# Plans and billing Source: https://webglazer.com/docs/account/plans-and-billing WebGlazer plans and prices, the 7-day free trial, upgrades and downgrades, invoices and how to cancel. This page explains what each WebGlazer plan includes, how the free trial works, and how to change, pay for or cancel your plan in the dashboard. > [!NOTE] > > Prices can change. The current prices are always on [webglazer.com/pricing](https://webglazer.com/pricing) for brands and [webglazer.com/pricing-agencies](https://webglazer.com/pricing-agencies) for agencies. Prices exclude VAT. If you buy as a private person, you pay the price plus VAT. A business in another EU country that enters a valid VAT number at checkout is reverse-charged and pays the listed price. There are no setup fees. ## Plans for brands | Plan | Monthly | Per month, billed annually | Projects | Daily tracked prompts | AI engines | | - | - | - | - | - | - | | Basic | €69 / \$79 | €59 / \$69 | 1 | 50 | Choose 3 | | Standard | €209 / \$239 | €179 / \$209 | 2 | 100 | All 5 | | Advanced | €419 / \$489 | €359 / \$419 | 5 | 200 | All 5 | | Enterprise | Custom | Custom | Custom | Custom | Up to 5 | ### Compare brand plans | | Basic | Standard | Advanced | | - | - | - | - | | Projects | 1 | 2 | 5 | | Daily tracked prompts | 50 | 100 | 200 | | Prompt slider goes up to | 200 | 250 | 350 | | AI engines | Choose 3 of 5 | All 5 | All 5 | | [Products](/docs/dashboard/products) (Shopify) | Free trial only | Included | Included | | Team seats | Unlimited | Unlimited | Unlimited | | Data exports and daily tracking | Included | Included | Included | | Support | Standard | Priority | Priority | | 7-day free trial | Yes | No | No | Enterprise plans are set up per customer; book a call to get a quote. [Products](/docs/dashboard/products) (product visibility for your Shopify store) is included in Standard, Advanced and every agency plan, and during the Basic free trial. A paid Basic plan does not include it. Every brand plan except Enterprise has a prompt slider, so you can track more prompts without moving up a tier. It moves in steps of 50, and the price updates as you slide. See the slider on [webglazer.com/pricing](https://webglazer.com/pricing). ## Plans for agencies Agency plans use credits instead of a daily prompt limit. One credit covers one prompt answered by one AI engine on one day. Credits are shared across your active client projects and pitch projects. | Plan | Monthly | Per month, billed annually | Active projects | Active pitch projects | Credits per month | | - | - | - | - | - | - | | Agency Basic | €189 / \$219 | €159 / \$189 | 5 | 3 | 14,400 | | Agency Plus | €399 / \$459 | €339 / \$389 | 10 | 5 | 30,600 | | Agency Max | €639 / \$739 | €539 / \$629 | 25 | 7 | 70,200 | | Enterprise | Custom | Custom | Custom | Custom | Custom | Each agency plan also has an Expanded option with more credits for a higher price: 21,600 credits a month on Agency Basic, 43,200 on Agency Plus and 91,800 on Agency Max. Every agency plan allows up to 60 prompts per pitch project and lets you choose 1 to 5 AI engines per project. Your daily credits are your monthly credits divided by 31. See [Agencies](/docs/setup/agencies#daily-credits). ## Free trial The Basic plan comes with a 7-day free trial for accounts that have never had a subscription or a trial. When the trial is available to you, the Basic plan shows **Start 7-day free trial**. If you are not eligible, for example after cancelling an earlier plan, the plan buttons read **Subscribe**. * You enter a payment card at checkout. You are not charged during the 7 days. * On day 7 the trial turns into a paid Basic plan by itself and tracking keeps running. * To stop before you are charged, cancel during the trial. See [Cancel your plan](#cancel-your-plan). Other plans and agency plans start without a trial. ## Where to manage billing Open **Settings** > **Billing & plan**. You can also choose **Billing** from the account menu at the bottom of the sidebar. Only a Company Admin can change the plan or billing. Other members see "Billing is managed by your company admin." The page shows: * **Your plan**: your current plan and its status, such as **Renews on** a date, **Trial ends** on a date, **Cancels on** a date, or **Payment failed**. It also shows **Next billing date**, or **Access until** if you have cancelled. * **Add more prompts** (brand plans) or **Add more credits** (agency plans): a slider to buy extra capacity on your current plan. * **Choose Your Plan**: the plan cards, with a **Brands** / **Agencies** switch and a **Monthly** / **Annual** switch on each card. * **Billing**: **Payment method** to update your card and **Invoices** to view and download past invoices. ## Start a plan **1. Open Billing & plan** Go to **Settings** > **Billing & plan**. **2. Pick a billing period** Use the **Monthly** / **Annual** switch on the plan card. Annual shows how much you save per year. **3. Click the plan's button** Click **Start 7-day free trial**, **Start with** plus the plan name, or **Subscribe**, whichever the card shows. You go to Stripe to pay. Stripe asks for your billing address and lets you enter a VAT or tax ID. For Enterprise, click **Book a call**. ## Upgrade or switch plans On a different plan's card, click **Switch to** plus the plan name. Stripe's billing page opens at the plan change and handles the price difference. ## Downgrade When you move to a smaller plan, the dashboard shows **Schedule Downgrade**. The new plan starts at the end of your billing period, and your current plan stays until then. Click **Schedule downgrade** to confirm. > [!WARNING] > > If you have more Active prompts than the new plan allows on the change date, tracking pauses until you archive the extra prompts. Archive them on **AI Visibility** > **Prompts** before the date to keep tracking. If you already have a downgrade scheduled and want a change to start today instead, the dashboard shows **Change plan now**. The new plan starts at once, replaces the scheduled change, and Stripe charges the difference for the rest of the billing period. ## Add more prompts or credits On **Billing & plan**, use **Add more prompts** (brand plans, **Extra prompts per day**) or **Add more credits** (agency plans, **Extra credits per month**). Move the slider, then confirm. A raise shows **Amount due now** and the new total, and you pay with **Confirm and pay**. Lowering or removing the add-on takes effect at the end of your billing period. ## Invoices and payment method Under **Billing**, click **Invoices** to view and download past invoices, or **Payment method** to update your card. Both open Stripe's secure billing page. If a payment fails, the page shows **Payment failed** and an **Update payment** button. ## Cancel your plan **1. Open your plan** Go to **Settings** > **Billing & plan**. On your current plan's card, click **Manage subscription**. **2. Cancel in Stripe** Stripe's billing page opens. Cancel your subscription there. **3. Check the end date** Back in WebGlazer, your plan shows **Cancels on** a date, and **Access until** that date. You keep access until then. Changed your mind? Click **Resume subscription** before the end date. To delete your account and data, email [support@webglazer.com](mailto:support@webglazer.com). # Settings Source: https://webglazer.com/docs/account/settings What each section of WebGlazer Settings does and where to change your account, billing, brand profile, projects and integrations. Settings is where you change how WebGlazer tracks your brand and manage your account. This page explains each section and what you can change there. ## Open Settings Click your name at the bottom of the sidebar and choose **Settings**. The sidebar switches to the Settings menu. Click **Back to app** to return to the page you came from. Settings has two groups: | Group | Section | What it is for | | - | - | - | | Account | **Account** | Your credit balance (agency accounts) | | Account | **Billing & plan** | Your plan, add-ons, payment method and invoices | | Workspace | **Brand profile** | What AI should know about the current brand | | Workspace | **Brands** | Your brand and the competitors you track | | Workspace | **Projects** | Every brand you track, its engines, time zone and prompt use | | Workspace | **Members** | Your team and their roles | | Workspace | **Integrations** | Google Search Console and your AI agent (MCP) | **Brand profile** and **Brands** apply to the brand selected in the sidebar's brand switcher. **Projects**, **Members** and **Integrations** cover the whole company. Some changes need the **Company Admin** role. If you cannot edit a field, the page says so. ## Account Agency accounts see their credit balance here. To delete your account or company, email [support@webglazer.com](mailto:support@webglazer.com). There is no delete button in the app. ## Billing & plan Shows your plan and its renewal or trial end date, lets you add prompts or credits, change or cancel your plan, update your card and download invoices. Only a Company Admin can make changes. See [Plans and billing](/docs/account/plans-and-billing) for every step. ## Brand profile **Brand Profile** holds what the AI knows about your brand. Your next run uses these answers. * **Logo**: click **Upload logo** or **Change logo**. Logos must be images under 2 MB. * **Market and language**: the market and language your buyers search in. Click **Edit** (or **Choose** / **Confirm** if it is not set yet) to change it. New AI prompts and your next run use this. Existing prompts stay as they are. Only a Company Admin can change it. The right side of the page holds your business details: * **Description** and **Category**. * **What you sell**: your products or services. Use **Add a product** to add one. * **Who you sell to**: your buyers. * **Question mix**: sliders for **Learning**, **Comparing** and **Buying**. They steer prompts you add later on the Prompts page. They never change prompts you already track. The three values must add up to 100%. After you edit, a bar appears. Click **Save changes**. Competitors are not edited here. Use **Brands**. ## Brands Your brand and the competitors you compare against, plus **Brand suggestions**: brands AI named in your answers that you do not track yet. The **Brands** menu item shows how many suggestions are waiting. See [Competitors](/docs/setup/competitors) and [Brands](/docs/setup/brands). ## Projects A table of every brand (project) you track, with its market, active prompts against your plan limit, AI engines, **Frequency** and creation date. Click a project's name to switch to it. ![Settings, Projects section with prompt usage and the project table with its row menu open](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-settings-projects.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=de363b571b0b817aee974cc61c086e36) Illustration with sample data * Click the engines to open **Tracking setup**, turn engines on or off, and click **Update**. * The row menu has **Edit details** (market, language and **Time zone**), **Tracking setup**, **Export chats** (a CSV of the project's AI answers) and **Delete project**. * The time zone sets when daily runs happen and which date they count for. Only a Company Admin can change it. * Deleting a project asks you to type its domain to confirm. See [Engines and schedule](/docs/setup/engines-and-schedule) and [Brands](/docs/setup/brands). Agency accounts also see **Clients** and **Pitches** tabs here; see [Agencies](/docs/setup/agencies). ## Members Invite teammates, change their roles and remove access. See [Team](/docs/setup/team). ## Integrations **Integrations** connects the tools that feed your prompts and the AI clients that read your results. ![Settings, Integrations with Google Search Console and Your AI agent, Setup guide and Clients and tokens open](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-settings-ai-agent.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=163205b55408add6b6349b6814cb2b65) Illustration with sample data * **Google Search Console**: connect Search Console to compare Google search data with AI answers and to add prompts from your real search queries. See [Search Console](/docs/dashboard/search-console). * **Your AI agent**: connect an AI client such as Claude or Cursor to your WebGlazer data through the WebGlazer MCP server. See [MCP server](/docs/integrations/mcp). # Core concepts Source: https://webglazer.com/docs/concepts Plain definitions of brand, project, prompt, topic, AI engine, run, competitor and credits in WebGlazer. WebGlazer uses a small set of words across the dashboard. This page explains each one so you can read your results and set up tracking with confidence. ## Brand (project) A brand is one website you track, such as your company or one of your clients. In **Settings** the same thing is called a **Project**. In these docs, brand and project mean the same thing. Each brand has its own website, market and language, prompts, topics, competitors and results. Switch between brands with the brand switcher in the sidebar. Add another one with **Add new brand** (agencies see **Add new client**), if your plan has room. How many brands you can track depends on your plan. See [Plans and billing](/docs/account/plans-and-billing) and [Brands](/docs/setup/brands). ## Prompt A prompt is one question WebGlazer asks the AI engines on your behalf, written the way a buyer would ask it. For example: "What is the best project management tool for small agencies?" Prompts have two states: * **Active**: tracked in every run and counted against your plan's prompt limit. * **Archived**: no longer tracked. You can restore an archived prompt, or delete it for good. Most prompts do not name your brand, so they show whether AI suggests you on its own. Prompts that name your brand are branded prompts. Branded prompts count in some numbers and not in others: | Number | Branded prompts counted? | | - | - | | [Visibility](/docs/metrics/visibility), [Share of Voice](/docs/metrics/share-of-voice), [Mention order](/docs/metrics/mention-order) | No | | [Sentiment](/docs/metrics/sentiment) | Yes | | [Citations](/docs/metrics/citations): **Times linked** and **Reach** | Yes | | [Query Fanouts](/docs/dashboard/fan-out) | Yes | | Which sources are qualifying, and [Opportunities](/docs/dashboard/opportunities) | No | See [Prompts and topics](/docs/setup/prompts-and-topics). ## Topic A topic is a group of related prompts, such as one product line or one buyer need. Topics let you filter results and see which subjects you are strong or weak in. During setup WebGlazer suggests topics from your brand profile and writes prompts for each one. You can add, rename and remove topics later on the **AI Visibility** > **Prompts** page. Removing a topic keeps its prompts and their history. ## AI engine An AI engine is one AI assistant WebGlazer asks your prompts. The dashboard also calls them platforms or models. WebGlazer tracks five: * ChatGPT * Perplexity * Google AI Mode * Gemini * Copilot WebGlazer asks each AI engine your prompts every day and records the answers. Your plan sets how many engines you can track. See [Engines and schedule](/docs/setup/engines-and-schedule). ## Run A run is one round of asking every active prompt on every engine you track, then reading the answers. Each run is a snapshot of how AI talked about you that day. You do not start runs yourself: they happen on a daily schedule. **1. Your active prompts** Every prompt you track that is not archived. **2. Asked on every engine you track** Each prompt goes to each AI engine in your plan. **3. Each answer is read** WebGlazer records the brands named, their order, quotes about you and the pages cited. **4. Your metrics update** Visibility, Share of Voice, Mention order, Sentiment and Citations. * Your first run starts when you finish setup. * After that, each brand runs once a day, at the same local time. * Your first brand runs at the time of day you created it, in your browser's time zone. A brand you add later copies the run time and time zone of your newest other brand. * The dashboard does not show this time of day, and you cannot change it. You can change the time zone, per brand, in **Settings** > **Projects**, in the row menu under **Edit details**. Only a Company Admin can do this. The dashboard shows results by date, so you can compare days and see what changed. See [Engines and schedule](/docs/setup/engines-and-schedule). ## Metrics at a glance Your results come down to five metrics. Each one answers a different question about how AI talks about you. | Metric | Question it answers | Range | Main place to see it | | - | - | - | - | | [Visibility](/docs/metrics/visibility) | How often does AI name my brand? | 0 to 100%, higher is better | **AI Visibility** > **Visibility** | | [Share of Voice](/docs/metrics/share-of-voice) | Of the mentions of my brand and tracked competitors, how much is mine? | 0 to 100%, higher is better | **Overview** > **Rankings** | | [Mention order](/docs/metrics/mention-order) | How early in the answer does AI name me? | 1 and up, lower is better (1 = named first) | **Overview** > **Rankings** | | [Sentiment](/docs/metrics/sentiment) | How well does AI speak about me? | 0 to 100, higher is better | **AI Visibility** > **Sentiment** | | [Citations](/docs/metrics/citations) | Which websites does AI use, and do they mention me? | **Times linked** (a count) and **Reach** (0 to 100%) | **AI Visibility** > **Citations** | ## Competitor A competitor is another brand you want to compare yourself with. WebGlazer checks every answer for your brand and your competitors, so you can see who AI names instead of you. You add competitors during setup under **Who you compete with**, and later in **Settings** > **Brands**. WebGlazer also lists **Brand suggestions**: brands AI named in your answers that you do not track yet. Accept a suggestion to start tracking it. See [Competitors](/docs/setup/competitors). Only your own brand and the competitors you track count in competitor numbers: **Rankings**, Share of Voice, the brand lists on the **Prompts** and **Visibility** tabs, and topic rankings. A brand AI names that you do not track is left out of all of them, including the Share of Voice total. Accepting it from **Brand suggestions** is what makes it count. ## Credits (agency plans) Agency plans measure usage in credits instead of a prompt limit. One credit covers one prompt answered by one AI engine on one day. For example, 10 prompts on 3 engines use 30 credits a day. * Credits come as a monthly allowance shared across all your client and pitch projects. * Your daily credits are your monthly credits divided by 31, rounded down. See [Agencies](/docs/setup/agencies#daily-credits). * The sidebar shows **Credits** with what is left this month. * When you choose engines, the dashboard shows how many daily credits that choice uses. * If you run out, runs pause until your credits reset for the next month. Brand plans (Basic, Standard, Advanced) do not use credits. They have a daily prompt limit instead. See [Agencies](/docs/setup/agencies). ## Roles Each person in your company has a role: **Company Admin**, **Member** or **Viewer**. Only a Company Admin can change billing, the market and language, or the time zone. Viewers can read results but not change setup. See [Team](/docs/setup/team). ## Client and pitch projects (agency plans) On an agency plan, every project is a client or a pitch. A client project is tracked every day for as long as you work with that client. A pitch project is a short 7-day project for a prospect, which you can convert to a client when they sign. See [Agencies](/docs/setup/agencies). ## Next steps > [card] [Quickstart](/docs/quickstart) > > Go from sign up to your first results. > [card] [Visibility](/docs/metrics/visibility) > > How often AI names your brand, and how it is calculated. > [card] [Prompts and topics](/docs/setup/prompts-and-topics) > > Add, edit and archive the questions WebGlazer asks. > [card] [Competitors](/docs/setup/competitors) > > Choose the brands you compare yourself with. # Citations tab Source: https://webglazer.com/docs/dashboard/citations The AI Visibility > Citations tab: the websites AI cites for your prompts, which engines cite them, and whether their pages mention you. The **Citations** tab lists the websites AI links to when it answers your prompts. "If these sites talk about you, AI talks about you." Use it to pick the sites worth getting featured on. Open it from **AI Visibility** > **Citations**. ![Citations tab with the category bar and the source table](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/wg-citations-2026-11.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=1debe37c8f082825e8738980f4897c9f) Sample data based on a linear.app project ## Qualifying or All sources A toggle in the top corner switches the list: * **Qualifying** (the default): sources cited by 3 or more engines across 3 or more prompts. The title reads, for example, "24 sources cited by 3+ engines across 3+ prompts". A source cited once by one engine is not a target, so it stays off this list. * **All sources**: every cited source, including the ones cited once by a single engine. The title reads, for example, "130 sources cited at least once". If no source clears the bar yet, you see "No sources cited by 3+ engines across 3+ prompts yet" with a hint to switch to **All sources**. ## Category bar A bar above the table splits all citations by kind of site, for example **News & blogs**, **Comparison sites** or **Your site**, each with its share. It shows up to 5 of the biggest categories and rolls the rest into **Other**. Each row's dot in the table uses the same colors. ## The source table | Column | What it shows | | - | - | | **#** | The source's rank among qualifying sources, most linked first. A dash for a source under the bar. | | **Site** | The website host. Click it to open the source detail. | | **Category** | What kind of site it is, as a colored dot. Hover for the name. | | **Cited by** | Which AI engines cited this source. | | **Times linked** | How many answers linked to this source, as a ring split by engine. Hover for the count per engine. | | **Reach** | How many of your topics this source is cited across, as five bars. Hover for "Cited across 3 of 10 topics · 30%". | | **Mentioned?** | Whether the source's checked pages name you, as "found / checked". **Your site** for your own domain. | How each column is counted is on the [Citations](/docs/metrics/citations) metric page. * Click a column header to sort. Each click cycles ascending, descending, off, and you can stack sorts. The **#** rank stays with each source. * **Export** downloads the table as CSV or JSON with Site, Category, Cited by, Times linked, Reach (topics) and Mentions you. * All numbers cover every run in the selected period, with the topic, tag and platform filters applied. ## Source detail Click a site to open its detail. The header reads, for example, "The 5 pages AI opened here when answering about you". * **Coverage**: a pill and a line such as "Mentioned on 2 of 5 checked pages". For your own site it says there is no mention to check. * **Named alongside**: how many AI answers named you next to this source. This counts the answers, not the page itself. * **Pages**: the checked pages, with **Opens** (how often AI opened each). **Show all N pages** lists every cited page with a **Mentioned?** verdict per page, and a search box. * **Check more pages**: shows in the **Show all N pages** view. It reads pages WebGlazer could not open on its own, so their mention verdict can be filled in. Each check reads up to 20 pages. **Check more pages** limits: * Each person can run 20 checks a day. The count resets at midnight UTC. It is the same on every plan and does not use your prompts or credits. * At the limit you see "Daily limit reached. You've used all of today's page checks, so try again tomorrow." * Members and Company Admins can run checks. Viewers cannot. If the tab says "No citations yet", it fills in after your next daily run. # Query Fanouts Source: https://webglazer.com/docs/dashboard/fan-out The AI Visibility > Query Fanouts tab: the web searches AI engines ran before answering your prompts, and whether those searches lead to your site. Before an AI engine answers, it often runs its own web searches first. These are called fan-out searches. The **Query Fanouts** tab lists them, so you can write pages that answer the exact searches AI runs. Open it from **AI Visibility** > **Query Fanouts**. ![Query Fanouts tab with the summary, What AI Adds to Your Prompts and What AI searched for](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-query-fanouts.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=7cc88b29ec9261ecebe1998a5420a1ef) Illustration with sample data ## Summary The card at the top shows how many different searches AI ran in the period ("searches AI ran") and a short read of how many of them lead to your site: * "None of them point to your site yet. Right now AI learns about your world from everyone except you." when no search led to your site. * "AI already points to your site on N. The rest are wide open." when under 34% of the searches did. * "AI already points to your site on N of them (X%). It clearly sees you as worth quoting." at 34% or more. How it is counted: * Searches that differ only in capitals, spacing or a trailing "?", "." or "!" count as one search. * A search counts as pointing to your site when at least one answer that ran it cited your own domain. * Coverage = searches that point to your site ÷ all different searches × 100, as a whole percent. > [!NOTE] > > **Example with made-up numbers.** AI ran 40 different searches in the period. Answers that ran 9 of them cited your site. Coverage = 9 ÷ 40 × 100 = 22.5%, shown as 23%, so the card reads "AI already points to your site on 9. The rest are wide open." If your project has no own domain set, the card shows only the count, with no coverage read. ## What AI Adds to Your Prompts "When AI searches, it throws in extra words you never wrote. Here are the ones it uses most. Put them on a page so AI finds you." The top 16 words that appear in AI's searches but in none of your tracked prompts, each with how many times a search containing it ran. Short filler words (such as "the", "for", "how") and words under 3 letters are left out. > [!NOTE] > > **Example with made-up numbers.** You track "best CRM for startups". AI searches "best CRM for startups 2026 pricing" 5 times. "2026" is not counted, because a word must start with a letter. "pricing" gets 5. ## What AI searched for "These are the exact searches AI ran before it answered. Answer one of them on your site and AI can quote you back." | Column | What it shows | | - | - | | **#** | The row's place in the list. | | **Query** | The search text. A green dot means you are cited when AI runs this search. An amber dot means you are not, a gap to win. No dot when your project has no own domain to check. | | **Platforms** | The engines that ran this search. | | **Times** | How many times the search ran in the period. | * The default order is most-run first. * Click **Query**, **Platforms** or **Times** to sort. Each click cycles ascending, descending, off. * **Export** downloads the list as CSV or JSON with Query, Platforms and Times. ### Open the answer behind a search Click a row to open the prompt detail sheet on **Raw data** for the engine that ran the search. If several engines ran it, the engine logos spread apart: click the one you want. The sheet works for prompts from older runs in the period too. See [Prompts tab](/docs/dashboard/prompts#the-prompt-detail-sheet). ## What counts Every successful answer from the run's selected engines, over every run in the selected period, with the topic, tag and platform filters applied. Branded prompts are included. If the tab says "No fan-out searches yet", no answer in the period ran a search. They appear after your next daily run. Not every engine shares its searches. Google AI Mode never does, so it never appears on this tab, and filtering to Google AI Mode alone shows an empty list. ChatGPT shares its searches on some answers and not others. # Opportunities tab Source: https://webglazer.com/docs/dashboard/opportunities The AI Visibility > Opportunities tab: sites AI cites for your prompts where your brand is not mentioned, with priority and the questions they feed. The **Opportunities** tab turns your citation data into a pick-list: "Get on the sources AI reads". Each card is a site AI cites to answer your questions where you are not mentioned. Get featured there and you can show up in those answers. Open it from **AI Visibility** > **Opportunities**. ![Opportunities tab with site cards, each with Add to to-do and Dismiss](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/wg-opportunities-2026-11.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=2eea5a994c27901798426f05cb503d86) Sample data based on a linear.app project ## Which sites become an opportunity A cited site shows up here only when all of these are true: * It is a qualifying source: cited by 3 or more engines across 3 or more non-branded prompts in the period. See [Citations](/docs/metrics/citations). * It is not your own site and not the site of a competitor you declared. * At least one of its pages has been checked, and no checked page names you. Only non-branded prompts count, because a branded prompt already names you. The page filters and date range apply, the same as on the **Citations** tab. ## What a card shows * **The site** (its host name). * **Priority**: **High priority** when the site is cited for 10 or more of your questions, **Medium priority** otherwise. * **Why**: a sentence such as "AI cites example.com to answer 4 of your 25 questions, and you're not on it. Get listed here and you're in the answer for all 4." * **The questions** it feeds: the first 3, then "+N more". * **Mentioned here instead**: competitors you declared that AI named in the answers citing this site. * "Back on your list: a newer run still shows this gap." when you marked it done and a later run cites the site again while it still does not name you. ## Acting on a card * **Add to to-do**: moves the card to your **To-do list** (in the sidebar), where you mark it **Done** or **Remove** it. See [To-do list](/docs/dashboard/todo). * **Dismiss**: hides the card. To bring a dismissed card back, open **To-do list** > **History** and click **Reopen**. This tab shows open cards only. When everything is picked or dismissed, it says "Nothing open. Check your To-do list." With no cards at all, it says "No opportunities yet". Your choices (to-do, done, dismissed) are your own. Other users of the project keep their own list. # Overview Source: https://webglazer.com/docs/dashboard/overview What the Overview page shows: the brand visibility chart and the Rankings table for your brand and the competitors you track. The **Overview** page is the first page you land on. It answers one question: how visible are you in AI answers compared with the other brands in your market? Open it from **Overview** at the top of the sidebar. ## Filter bar The bar at the top sets the scope for every number on the page. * The date pill: **Last 7 Days** (the default), **Last 30 Days**, **Last 90 Days**, **Last 6 Months**, **Last 12 Months**, **All time**, or **Custom range** to pick days on a calendar. Every number is calculated over all runs in the range. * **Topics**: limit the page to one topic, or **All topics**. * **Tags**: limit the page to prompt types (search intents), for example **Informational**, **Consideration** or **Transactional**. Older runs can also carry older tags, such as **Commercial**, **Navigational**, **Comparison** or **Post-purchase**, and these show in the picker too. * **Platforms**: limit the page to one or more AI engines. * **Reset**: shows once you change anything, and puts every filter back to its default. The branded topic and the **Branded** tag are hidden on this page. Branded prompts already name you in the question, so they cannot produce a visibility rate. If no run falls inside the filters, the page shows "No runs match your filters." with a way to reset them. ## Visibility Over Time A chart of how often AI mentions each of the top 10 brands from **Rankings**. It has two views, switched with the toggle in the card's top corner: ![Visibility Over Time chart in the By date view, one line per brand over 90 days](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/wg-visibility-over-time-2026-11.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=d67025b834d19e09187f28b96f9d886d) Sample data based on a linear.app project * **By brand** (the default): one bar per brand, showing its [Visibility](/docs/metrics/visibility) for the whole period. This answers "who is ahead right now". * **By date**: one line per brand across the period. This answers "how has it moved". It needs runs on at least two dates; otherwise it says "Not enough runs in this period to chart a trend". Click brand chips in the legend to show only those brands. Click again to remove them. Use **Export** to download the chart data as CSV or JSON: one row per date, one column per brand. ## Rankings "Brands ranked by how often AI mentions them." One row per brand: your own brand and the competitors you track. A brand AI names that you do not track is not listed. Accept it in **Settings** > **Brands** to add it. See [Competitors](/docs/setup/competitors). ![Rankings table with seven brands and their Visibility, SOV, Sentiment and Mention order](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-rankings.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=3c3667426ee4999b5d95971ef914aa3d) Illustration with sample data | Column | What it shows | | - | - | | **#** | The brand's rank, with a chip showing places gained or lost against the previous period. | | **Brand** | The brand name and logo. | | **Visibility** | Share of answers that name the brand. See [Visibility](/docs/metrics/visibility). | | **SOV** | The brand's share of all brand mentions. See [Share of Voice](/docs/metrics/share-of-voice). | | **Sentiment** | How well AI talks about the brand, 0 to 100. Shown for your own brand only. See [Sentiment](/docs/metrics/sentiment). | | **Mention order** | Average place where AI first names the brand in the answer. 1 = named first. See [Mention order](/docs/metrics/mention-order). | ### How brands are ranked The default order is set by WebGlazer, not by the browser: 1. Higher Visibility first. 2. On a tie, higher Share of Voice. 3. Then more answers naming the brand. 4. Then the better (lower) Mention order. 5. On a full tie, your own brand first, then by name. Without topic, tag or platform filters, a competitor needs at least 3 answers naming it to get a rank. Your own brand always gets one. With a filter on, every tracked brand AI names ranks, because the slice is small. The table lists at most 50 competitors, plus your own brand and every competitor you declared, wherever they rank. ### Sorting, change chips and export * Click a column header to sort by it. Each click cycles ascending, descending, off. Sort a second column to break ties in the first. The **#** rank stays with each brand. * Change chips compare with the previous period of the same length, for example the 7 days before the current 7 days. **All time** has no chips. If the previous period has no data, you see a "no comparison" note instead. * Use **Export** to download the table as CSV or JSON. * With more rows than one page holds, use the pager under the table. ## While a run is still being read When a run's answers are in but WebGlazer is still reading them for brands and sentiment, a banner says "Analysing competitors & sentiment". Share of voice, rankings and sentiment can still change until it finishes. # Platforms tab Source: https://webglazer.com/docs/dashboard/platforms The AI Visibility > Platforms tab: each AI engine side by side with your visibility, mention order, sentiment and citations. The **Platforms** tab compares the AI engines you track. It answers: which engine mentions me most, which names me earliest, which likes me best, and which reads the most sources? Open it from **AI Visibility** > **Platforms**. ![Platforms tab: How each engine treats you, one row per AI engine](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/wg-engines-2026-11.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=660150028bdb4557720396c078390fed) Sample data based on a linear.app project ## How each engine treats you "Every AI engine side by side: how often it mentions you, how early it names you when it does, how it feels about you, and how many sources it pulls." One row per engine your project tracks. | Column | What it shows | | - | - | | **Engine** | The engine's name and logo, for example ChatGPT or Google AI Mode. | | **Visibility** | Share of this engine's answers that name you, with a bar. See [Visibility](/docs/metrics/visibility). | | **Mention order** | Average place where this engine first names you in its answers. See [Mention order](/docs/metrics/mention-order). | | **Sentiment** | This engine's sentiment score for you, 0 to 100, with a bar split into positive, neutral and negative answers. See [Sentiment](/docs/metrics/sentiment). | | **Citations** | Times linked: how many times this engine's answers linked to a source. See [Citations](/docs/metrics/citations). | ## How the numbers work * Each engine stands alone. Its visibility divides only by the answers that engine returned successfully. An engine that answered 1 prompt and named you reads 100%. * Visibility and mention order count non-branded prompts only. Sentiment and **Citations** include branded prompts. So an engine can have a sentiment score while its visibility is empty. * All numbers cover every run in the selected date range, with the topic and tag filters applied. * The default order is highest visibility first. Engines with no visibility rate go last. * Click **Visibility**, **Mention order**, **Sentiment** or **Citations** to sort. Each click cycles ascending, descending, off, and you can stack sorts. Engines with no data always sort to the bottom. * A dash means the engine has no data for that column in the view. Use the **Platforms** filter in the filter bar to hide engines you do not want to compare. If the tab says "No platform data yet", it fills in after your next daily run. "No runs match your filters." means no run falls inside the filters. # Products Source: https://webglazer.com/docs/dashboard/products How to connect a Shopify store to WebGlazer and read product-level AI visibility: which of your products AI shopping answers show, how often, and next to whom. The **Products** page shows AI visibility for each product in your Shopify store. It tells you which products appear in the product cards of AI shopping answers, how often they come first, and which competing products and shops show up beside them. ![Products page for a connected store, on the Products tab with categories and the product table](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-products.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=2f1ccbf4c741cd51cacb017ae14e34a4) Illustration with sample data ## Who has Products Products is included in the Standard, Advanced and Agency plans, and in the Basic trial. A paid Basic plan has no Products. Without access there is no **Products** item in the sidebar, and a connected AI agent cannot read product data either. If you lose access, your imported data is kept and shown again when access returns. ## Connect your Shopify store **1. Open Products** Click **Products** in the sidebar, right below **AI Visibility**. **2. Enter the store address** Under **Connect your Shopify store**, check the store address. It starts filled with your brand's website. **3. Connect** Click **Connect store**. WebGlazer reads the store's public product feed, so there is nothing to install in Shopify. The first import shows its progress (**Waiting to start**, **Checking the store**, **Reading products**) and usually takes under a minute. You can leave the page while it runs. Only a **Company Admin** can connect or disconnect the store. Everyone else sees a note that a Company Admin can connect it. Limits: * One store per brand. * The first 1,000 products in the store are tracked. Click **View untracked products** to see the ones that are not. * Password-protected stores and stores with bot protection cannot be imported. The store imports once when you connect it. To import it again, disconnect and reconnect it. ## Where product numbers come from WebGlazer reads the shopping product cards that AI engines show when they answer your prompts, and matches each card to a product in your catalog. Product numbers update after each daily run. If none of your prompts got an answer with product cards, the page says **No product cards yet**. ## Overview tab **Products** > **Overview** sums up the period chosen in the filter bar: * **Brand Visibility**: how often each brand shows up in the product cards of your shopping answers, over time, with a **Top Brands** table. Your own brands are tagged **You**. * **Performance by Visibility**: your products AI shows most and least. * **Shopping Queries**: the shopping searches AI ran to pick the products in its answers. * **Query Fanouts**: the web searches AI ran while writing your shopping answers. * **Merchants**: the shops AI points shoppers to in its product cards. * **Sources**: the websites AI read while answering your shopping prompts. Lists have **Top**, **Trending**, **Losing** and **New** tabs. A category pill on the filter bar narrows the Overview to one product category. ## Products tab **Products** > **Products** lists every tracked product with: * **Visibility**: the share of shopping answers to this product's prompts that showed it. * **Win rate**: the share of its appearances where the product was the first card. * **Position**: its average card position. Lower is better. * **Rating**, **Catalog price** and **Mentioned price** (the price AI quoted). * **Merchants**, **Competing brands** and **Competing products**. * **Top shopping queries**, **Top query fanouts** and **Appearances**. Sort by any column, search by title, and export the list as CSV or JSON. **Categories** on the left filter by Shopify product type and show that category's visibility. Products with no type are under **Uncategorised**. ## A single product Click a product to open its page. It shows: * How the product was matched, and a link to it in your store. * **Visibility & Win Rate** over time and **Competing Products**. * **Position vs Competitors**: where this product and its most frequent competitors sit in the product cards. * **Merchants**, **Sources**, **Shopping Queries** and **Query Fanouts** for answers that showed this product. * **Prompts & Chats**: the prompts and answers this product appears in, and where it came first. ## Disconnect the store Click **Disconnect** in the store header and confirm. Product tracking stops. The imported catalog is kept hidden and past product results stay. Reconnecting the same store brings the catalog back and imports it again. # Prompts tab Source: https://webglazer.com/docs/dashboard/prompts The AI Visibility > Prompts tab: every tracked prompt with its visibility, mention order, volume and brands, and the prompt detail sheet with the raw AI answers. The **Prompts** tab lists every question you track and how AI answered it. Use it to find the prompts where you are missing, then open one to read the exact answers. Open it from **AI Visibility** > **Prompts**. ![Prompts tab with the topic rail, Active and Archived tabs, and the prompts table](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-prompts-list.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=e08276b300c0a50a1f32f78eb5d641c0) Illustration with sample data ## Your numbers above the table A line above the table shows your own brand across the prompts in view: **Share of voice**, **Visibility**, **Sentiment** (with a dot in the band color) and **Position**. They follow the page filters and date range. See [Share of Voice](/docs/metrics/share-of-voice), [Visibility](/docs/metrics/visibility), [Sentiment](/docs/metrics/sentiment) and [Mention order](/docs/metrics/mention-order) (shown here as **Position**). ## Topic rail On wide screens a rail on the left lists **All topics** and each topic with its number of active prompts. Click a topic to filter the table, and the rest of **AI Visibility**, to it. On smaller screens, editors get a **Manage topics** button instead. Adding, renaming and archiving topics is covered in [Prompts and topics](/docs/setup/prompts-and-topics). ## Active and Archived Two tabs sit above the table: * **Active**: the prompts WebGlazer runs. This is the default. * **Archived**: prompts you stopped tracking. Their last results stay visible and you can **Restore** them. Next to the tabs you see your prompt allowance as "used / allowance" and the **Add prompts** button. When you reach the limit, you get **Upgrade plan** and **Manage prompts** instead. ## The table Use **Search prompts** to filter rows by text. | Column | What it shows | | - | - | | **Prompts** | The question. The ring next to it is your [Visibility](/docs/metrics/visibility) for that prompt, as a whole percent. | | **Mention order** | Average place where AI first names your brand in the answers to this prompt. 1 = named first. See [Mention order](/docs/metrics/mention-order). | | **Volume** | The monthly Google search volume for the closest matching keyword, found when WebGlazer wrote the prompt. Prompts you write yourself show "No keyword data". | | **Brands** | Your brand and the competitors you track that AI named when answering this prompt. Hover to spread the logos. | | **Mentioned?** | **Yes** if any answer in the period named you, **No** if none did, **Unknown** if there is no usable answer yet. | | **Type** | The prompt's search intent as a colored dot, for example **Informational**, **Consideration**, **Transactional** or **Branded**. Hover for its meaning. | All numbers cover every run in the selected date range, with the filters applied. * Click a column header to sort. Each click cycles ascending, descending, off. * A prompt that has never run reads **Not run yet**. A prompt with no answer in the selected range reads **No result in this period**. * **Branded** prompts get a **Mentioned?** verdict but no visibility ring and no mention order, because they name you in the question. * Editors can select rows to archive them in bulk, and use each row's menu for **Edit prompt** and **Archive prompt**. * **Export** downloads every AI answer in the period and filters as CSV or JSON, one row per answer, with columns such as Prompt, Type, Topic, Model, Ran at, Brand mentioned, Sentiment, Competitors, Citation URLs, Fan-out queries and Answer. ## The prompt detail sheet Click any prompt row to open its detail sheet. It shows the prompt text, its type and its topic at the top, and has three tabs: **Insights**, **Raw data** and **Sources**. You can also reach the sheet from other tabs: * **Sentiment** tab: click a quote. The sheet opens on **Raw data** for that engine, with the quote highlighted. * **Query Fanouts** tab: click a search. The sheet opens on **Raw data** for the engine that ran it. * A link in a report opens the prompt on that engine's raw answer. The sheet has its own date pill. On **Insights** and **Sources** a models pill picks which engines the numbers cover (**All models selected** by default). On **Raw data** an engine picker picks whose answer you read. ### Insights * Three cards: **Mention order**, **Sentiment** and **Volume**. * **Visibility Over Time**: your visibility on this prompt across the period. A branded prompt shows "Prompts that name your brand get a yes/no mention, not a visibility score." instead. * **Top Brands**: your brand and the tracked competitors named for this prompt, with **Visibility** (share of this prompt's answers where the brand showed up), **Sentiment**, **Mention order** and **Sources** (how many AI engines named the brand for this prompt). ### Raw data ![Prompt detail on the Raw data tab, showing ChatGPT's answer and the searches it ran](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-prompt-detail-raw-data.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=32112b0727bdd569cd575bc07737165b) Illustration with sample data The full answer one engine gave, shown as a chat: your prompt, then the answer. It shows the newest successful answer from that engine in the period. Use the copy button to copy the answer text. Under the answer you may see what else the engine returned: **Searches the Engine Ran** (its fan-out searches), **Sponsored**, **Products**, **Places** and **Videos**. A section shows only when the engine returned it. ### Sources "Pages the AI answers cited for this prompt, across every engine. A source cited by more engines is more worth your attention." Each row shows the page title and site, its **Category** and **Mentions** (how many times it was cited). Click a row to open the page. # Reports Source: https://webglazer.com/docs/dashboard/reports Where to find the reports the WebGlazer team prepares for your brand, how to open and download them, and what a downloaded report contains. The **Reports** page holds every report the WebGlazer team has prepared for the brand you are viewing. Open a report in the dashboard to follow its links to the underlying answers, or download it to share outside WebGlazer. ## Find your reports Click **Reports** in the sidebar. The list shows each report with its name and **Date**, newest first. ![Reports list with each report's name, date and an Open or Download button](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-reports.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=f5d54f765c94a021be71047ff2a89032) Illustration with sample data Reports are prepared and added by the WebGlazer team. You cannot create one yourself. If the list says **No reports yet**, the team has not added one for this brand. Reports show up here once they are ready. Reports belong to a brand. Switch brands in the sidebar to see another brand's reports. ## Open a report Click **Open** on a report. It opens on its own page inside the dashboard. * Answers in the report link to the prompt they came from. Clicking one opens that prompt in **AI Visibility** > **Prompts** in a new tab, on the engine that gave the answer. * Click **Back to reports** to return to the list. Some reports are spreadsheet files. Those show **Download** in the list instead of **Open** and download the file directly. If a report cannot be shown, the page says **Report not available**. Go back to the list and try another report. ## Download a report On an open report, click **Download** to save it as an HTML file you can send to anyone. > [!NOTE] > > The downloaded file keeps the numbers but leaves out the AI answers and the links back into the dashboard. Those parts only work for people who can sign in to your WebGlazer account. # Search Console Source: https://webglazer.com/docs/dashboard/search-console How to connect Google Search Console to WebGlazer, what the Search Console page shows next to AI citations, and how to disconnect or remove it. The **Search Console** page puts your Google search data next to how AI answers cite the same pages. It shows which pages rank on Google but are ignored by AI, and which pages AI cites but Google barely shows. ![Search Console Overview with clicks, impressions, the chart and top queries and pages](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-search-console.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=12b60e50a6fe6ae8f42b1049b28e3895) Illustration with sample data ## Connect Google Search Console Only a **Company Admin** can connect. Anyone else with access to the brand sees the data once it is connected. **1. Open Search Console** Click **Search Console** in the sidebar. **2. Sign in with Google** Click **Connect with Google** and sign in with a Google account that has your website in Search Console. On Google's screen, tick "View Search Console data for your verified sites" before you continue. **3. Choose websites** In **Choose Websites to Track**, every website on the account starts ticked. Untick the ones you do not want, or click **Only this** next to one. Then click **Save N websites**. You can track up to 50 websites. WebGlazer has read-only access and never changes anything in Search Console. The connection belongs to the brand you are viewing. After you save, the **First Import** loads 16 months of Google data, newest days first. You can leave the page while it runs. After that, WebGlazer collects new days every day. ### If connecting fails * **Nothing was connected**: you declined access on Google's screen. Try again. * **Google didn't give access to Search Console**: the Search Console box was not ticked. Try again and tick it. * **Your Google Workspace admin has blocked this app**: ask your Workspace admin to allow it, or use a personal Google account that has the website. * **This Google account has no Search Console websites**: add the site in Search Console first, or click **Use a different account**. ## Filters Every tab shares the filters at the top: the website, the date range (**Last 7 days**, **Last 30 days**, **Last 3 months**, **Last 6 months**, **Last 12 months** or **Last 16 months**; default **Last 30 days**), countries and devices. Dates are Google's days in Pacific Time, as Google reports them. Google finalizes its numbers about 3 days later, so recent days are shaded on charts and may still change. Google counts each website separately, so numbers for two websites are never added together. ## Overview * **Clicks**, **Impressions**, **Click Rate** and **Average Position** for the period, with the change from the period before. * **Clicks and Impressions** over time, by **Day**, **Week** or **Month**. * **Top Queries** and **Top Pages**. Google hides rare queries, so top queries add up to less than the total. * **Clicks by Country** and **Clicks by Device**. ## Google vs AI This tab compares Google with the AI answers to your tracked prompts. It has two lists: * **Ranking on Google, Not Cited by AI**: pages with at least 1,000 impressions on Google's first page and fewer than 2 AI citations. These pages already earn Google traffic but AI does not use them as sources. * **Cited by AI, Barely on Google**: pages with at least 5 AI citations and under 100 impressions, or an average position worse than 20. Each page shows **AI retrievals**, **AI citations**, **Last crawled** and **Index status** from Google's page inspection, for example **Indexed as declared**, **Google picked another URL**, **Soft 404** or **Blocked by robots.txt**. This tab ignores the country and device filters. ## Queries and Pages **Queries** and **Pages** each show the top 5 over time, then a full table (**All Queries**, **All Pages**) you can sort and search. ## Settings tab A Company Admin manages the connection on **Search Console** > **Settings**: * **Google Account**: the connected account. **Sign in again** reconnects it. * **Tracked Websites**: tick or untick websites and click **Save changes**. * **Disconnect**: WebGlazer stops collecting Google data and stops accessing Search Console. The Google data already imported stays, and so do prompts you already added from Search Console. * **Remove Search Console**: deletes every Google number imported for this brand and disconnects the account. Type the name shown to confirm, then click **Remove and delete data**. To use it again you start from scratch. If Google access expires, a **Connection expired** banner appears and collection pauses. Click **Reconnect** to pick up where it stopped. ## Search Console in other places * **AI Visibility** > **Prompts** > **Add prompts** > **Google Search Console** turns question-style Google searches into prompt suggestions. See [Prompts and topics](/docs/setup/prompts-and-topics). * **Settings** > **Integrations** shows the connection status. It reads the same connection as the **Search Console** page. # Sentiment tab Source: https://webglazer.com/docs/dashboard/sentiment The AI Visibility > Sentiment tab: your sentiment score, the plain-language summary, and what AI says about you by category, down to the exact quotes. The **Sentiment** tab shows how AI talks about you and why. It starts with one score and a short read, then breaks the answers into categories so you can find the complaint worth fixing first. Open it from **AI Visibility** > **Sentiment**. ![Sentiment tab with the score ring, summary and the category table](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/wg-sentiment-2026-11.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=a3d047de5e6f2c301ac25324cb610bb5) Sample data based on a linear.app project ## Score and summary The ring on the left is your [Sentiment](/docs/metrics/sentiment) score, 0 to 100, colored by its band (positive, neutral or negative). It covers every run in the selected period, with the filters applied, and includes branded prompts. Next to it, WebGlazer writes a short read of where you stand: * **The headline** follows the trend first, then the band: "AI is warming up to you" or "AI has cooled on you a bit" when the score moved 3 points or more against the previous period of the same length; otherwise "AI speaks well of you", "AI is fairly neutral about you" or "AI is critical of you right now". * **The detail** names the topic AI praises most and the complaint that comes up most, for example "It speaks best about your support, but pricing keeps coming up." * **The action** points at that complaint: "That's the one to fix first.", or "New in this period, so worth a look." when it did not come up in the previous period. **See what they said** opens that category's quotes. * With fewer than 8 scored answers you also see "Still early, from just a handful of answers." ## What AI says about you, by category "Every answer sorted into fixed categories, complaints first. Click a category to read every topic behind its tone and who said each." Each row is one category: **Features**, **Ease of Use**, **Pricing & Value**, **Support & Service**, **Performance & Reliability**, **Quality**, **Integrations**, **Trust & Reputation**, **Onboarding** and **Other**. * **Category**: the kind of thing AI talked about. * **Tone**: the category's score from 0 (all bad) to 100 (all good) with a colored bar. Hover to see how many quotes were positive, neutral and negative. The score is (positive quotes + half of neutral quotes) ÷ all quotes × 100. Click the header to sort. Categories that are mostly negative come first, then the rest by how often they came up. **Other** is always last. **Export** downloads the categories as CSV or JSON with each category's sentiment, score, positive, neutral and negative mentions, total times and topics. ## Reading the quotes Click a category to open it. "Everything AI said in this category, every quote grouped by topic. The count is how many answers used that exact line." * Each topic inside the category is its own section, with its number of quotes. The most-said topic is open first. * Each quote row shows the **Engine** logos that said it, the **Quote** and its **Tone** (**Positive**, **Neutral** or **Negative**). Sort by **Engine** or **Tone**. * "Not in the shown answer text" marks a quote taken from the engine's full answer that is not in the part shown to you. * Click a quote to open the prompt detail sheet on **Raw data** for that engine, with the quote highlighted in the answer. See [Prompts tab](/docs/dashboard/prompts#the-prompt-detail-sheet). ## Empty states * "No sentiment yet" means no answer in the period has a sentiment score. It fills in after your next daily run. * "No scores yet. Check back after your next run." means categories exist but no score could be calculated yet. # To-do list Source: https://webglazer.com/docs/dashboard/todo How to work through the sources you picked from Opportunities on your To-do list, mark them done, and reopen them. The **To-do list** holds the sources you picked from **AI Visibility** > **Opportunities**: sites AI cites for your prompts where your brand is not mentioned. Use it to track each one until it is handled. Which sites become an opportunity, and what each card shows, is on the [Opportunities tab](/docs/dashboard/opportunities) page. ## Pick items for your list **1. Open Opportunities** Go to **AI Visibility** > **Opportunities**. If it says **No opportunities yet**, check again after your next daily run. **2. Choose what to work on** On each card, click **Add to to-do** to pick it, or **Dismiss** to hide it. ## Work through your To-do list ![To-do list page with the Connect your AI card and four picked sources](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-todo-list.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=25f3a83a075db703d2df6a4057488299) Illustration with sample data Click **To-do list** in the sidebar. **Your list** has two views: * **To-do**: what you picked. Click **Done** when you have handled a source, or **Remove** to send it back to Opportunities. * **History**: items marked done or dismissed. Click **Reopen** to put one back in Opportunities. The filter bar sets which runs the list is based on. If the list is empty, click **Pick tasks in Opportunities**. Your picks and statuses are personal. Teammates on the same brand keep their own list. Every role, Viewers included, can keep a list. ## When a done item comes back If a newer run still cites the source and it still does not name you after you marked it done, the item goes back to the **Opportunities** tab with the note "Back on your list: a newer run still shows this gap." Pick it again to put it back on your To-do list. Dismissed items stay dismissed. ## Use your AI agent The To-do list can be filled by an AI agent connected through the WebGlazer MCP server. Until you connect one, a **Connect your AI** card at the top explains how: 1. Generate a token in **Settings** > **Integrations** > **Your AI agent**. 2. Paste it into Claude Code, Cursor, or any MCP client. 3. Ask it for your opportunities. What it picks lands on this list. Click **Connect your AI** to open **Settings** > **Integrations**, or **Not now** to hide the card until your next browser session. The card disappears for good once you hold an active token. See [MCP server](/docs/integrations/mcp) for setup. # Topics tab Source: https://webglazer.com/docs/dashboard/topics The AI Visibility > Topics tab: how many prompts each topic holds, how visible you are in it, and how that moved. The **Topics** tab shows your visibility topic by topic. Use it to see which parts of your market AI already connects you with, and which ones it does not. Open it from **AI Visibility** > **Topics**. ![Topics Explorer with the tile map and the visibility trend per topic](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-topics.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=63f6b60f501799ce9c56a7c59507f434) Illustration with sample data ## Topics Explorer "Bigger tiles hold more prompts. Darker tiles are topics where AI mentions you more." ### Tile map * Each tile is one topic with active prompts. * Tile size is the number of active prompts in the topic. * Tile shade is your [Visibility](/docs/metrics/visibility) in that topic: the share of the topic's answers that name you, over the selected period and filters. Darker means higher. * The map draws the top 10 topics by prompt count. The label above it says how many it shows, for example "Top 10 of 14 topics by prompts". * On a phone, the map becomes a ranked list. ### Visibility trend Next to the map, one row per topic (the top 8 by prompt count) shows your visibility in that topic from day to day across the period. A day without a run breaks the line. A topic with no runs yet says "No runs yet". ### All topics With more than 10 topics, **See all topics (N)** swaps the trend for a table of every topic. Search it with the search box, and sort by **Topic**, **Prompts** or **Visibility**. **Show trend** switches back. ## Filtering by topic Click a tile, a trend row or a table row to set the page's topic filter. The whole of **AI Visibility** then shows only that topic. Click the selected topic again to clear it. Branded topics are left out of this tab, and the branded filter options are hidden, because branded answers have no visibility rate. If there are no topics with active prompts, the tab says "No topics with active prompts yet. Add prompts to a topic to see it here." # Visibility tab Source: https://webglazer.com/docs/dashboard/visibility The AI Visibility > Visibility tab: your Visibility Score, its trend across four metrics, and which brands lead each topic. The **Visibility** tab is the first tab of **AI Visibility**. It answers: how visible is my brand right now, how has that moved, and who leads each of my topics? Open it from **AI Visibility** in the sidebar. It opens on **Visibility** by default. ![AI Visibility page on the Visibility tab, with the Visibility Score chart and Visibility by Topic](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/wg-visibility-score-2026-11.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=95955c5a9a962be30c90e3ba692f2f6c) Sample data based on a linear.app project ## Tabs and filter bar **AI Visibility** has eight tabs: **Visibility**, **Prompts**, **Topics**, **Platforms**, **Sentiment**, **Citations**, **Query Fanouts** and **Opportunities**. The filter bar under them (date, **Topics**, **Tags**, **Platforms**, **Reset**) is shared by all tabs, so a selection carries over when you switch. See [Overview](/docs/dashboard/overview#filter-bar) for what each filter does. On this tab the branded topic and the **Branded** tag are hidden, because every number here is a rate that branded answers cannot produce. ## Visibility Score "How your visibility, share of voice, sentiment and mention order have moved across your runs. Tick a metric to plot its line." * **The big number** is your [Visibility](/docs/metrics/visibility) for the selected period and filters. * **The chip** next to it is the change in percentage points (pp) against the previous period of the same length. If there is nothing to compare with, it says so. * **Metric checkboxes**: **Visibility**, **Share of Voice**, **Sentiment** and **Mention order**. **Visibility** and **Share of Voice** are ticked at first. Tick a box to add its line. At least one line always stays on. A box only shows for a metric that has data in the period. * **The chart** has no Y axis on purpose. The four metrics use different scales, so each line is stretched to the full height to show its shape. Hover a point to read the real values. The **Mention order** line is flipped, so up always means better on every line. * **Points per period**: one point per day for ranges up to 35 days and for **All time**, one per week up to 120 days, one per month above that. Several runs on one day are pooled into one point. Days without a run are skipped. * With a single run in the period you see bars instead of lines. With none, "Not enough runs yet". * **Export** downloads the points as CSV or JSON with the columns Date, Visibility %, Share of Voice %, Sentiment and Mention order. ## Visibility by Topic "Every row is one of your topics. The logos are the brands AI brings up when people ask about it, from most mentioned to least." ![Visibility by Topic table, one row per topic with brand logos ranked #1 to #8](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/wg-visibility-by-topic-2026-11.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=80239eefa9ddcc17c7af9e5569a4aa58) Sample data based on a linear.app project * Each row is one topic. Columns **#1** to **#8** hold the brands ranked in that topic, best first. Your own brand's logo has a double outline so you can spot it down each column. * Hover a logo to see the brand name. * Topic rankings use the same order as **Rankings** on [Overview](/docs/dashboard/overview#how-brands-are-ranked), applied to that topic's answers. Every tracked brand AI names ranks here, with no minimum number of answers. Brands you do not track are left out. See [Competitors](/docs/setup/competitors). * If a brand is hidden by a filter, its slot stays empty instead of shifting the others left. * **Export** downloads one row per topic with the brand name in each slot. * On a phone, each topic collapses to a card with its top brands. ## Empty states * "No visibility data yet" means no answer in the period could be counted. It fills in after your next daily run. * "No runs match your filters." means no run falls inside the filters. Reset them. # FAQ Source: https://webglazer.com/docs/faq Answers to common questions about WebGlazer tracking, engines, data updates, competitors, billing and privacy. Short answers to the questions new WebGlazer users ask most. Each answer links to the page with the full details. > [accordion] Which AI engines does WebGlazer track? > > Five: ChatGPT, Perplexity, Google AI Mode, Gemini and Copilot. WebGlazer asks each engine your prompts every day and records the answers. The Basic plan tracks any 3 of them. Standard and Advanced track all 5. Agency plans choose 1 to 5. See [Engines and schedule](/docs/setup/engines-and-schedule). > [accordion] How often does my data update? > > Once a day. Each brand runs all its active prompts on all its engines once a day, at the same local time in the brand's time zone. Your first brand runs at the time of day you created it. Brands you add later copy the time of your newest other brand. A Company Admin can change the time zone in **Settings** > **Projects**, row menu, **Edit details**. See [Engines and schedule](/docs/setup/engines-and-schedule#how-often-runs-happen). > [accordion] Can I start an extra run myself? > > No. Runs happen on the daily schedule. There is no button in the dashboard to start an extra run. Changes you make to prompts, engines or your brand profile apply from the next daily run. If an empty screen says "Run your prompts", it fills in after your next daily run. > [accordion] Why do my numbers change from one run to the next? > > AI engines do not give the same answer every time. The same question can name different brands, in a different order, with different sources, on different days. Each run is a snapshot of what the engines said that day, so small day-to-day changes are normal. Look at the trend over several days rather than a single run. > > Your own changes also move the numbers: adding or archiving prompts, switching engines, or changing your market. Deleting an archived prompt removes all its results, so past numbers are recalculated without it. > [accordion] How long until I see my first results? > > Your first run starts as soon as you finish setup and usually takes about four to six minutes. The **Run Progress** popup shows one ring per AI engine filling up as answers arrive. When the run finishes, your results appear on the **Overview**. After that, new data arrives with each daily run. > [accordion] Can I track more than one country? > > Each brand tracks one market (country) and one language. You set them in **Settings** > **Brand profile** > **Market and language**. See [Settings](/docs/account/settings#brand-profile). > [accordion] What is the difference between a prompt and a topic? > > A prompt is one question WebGlazer asks the AI engines, such as "best CRM for small agencies". A topic is a group of related prompts, such as one product line. Topics let you filter results by subject. See [Core concepts](/docs/concepts). > [accordion] Why is a competitor missing from Rankings? > > There are two common reasons: > > * **You do not track it.** Only your own brand and the competitors you track appear in **Rankings** and count in Share of Voice. Add it in **Settings** > **Brands** with **Add brand**, or accept it from **Brand suggestions**. > * **AI named it in fewer than 3 answers.** With no topic, tag or platform filter on, a competitor needs at least 3 answers naming it to get a rank. > > See [Overview](/docs/dashboard/overview#how-brands-are-ranked) and [Competitors](/docs/setup/competitors). > [accordion] How do I add a competitor? > > Go to **Settings** > **Brands** and click **Add brand**. Enter the name and domain and click **Add brand** in the dialog. You can also accept a brand from **Brand suggestions**: brands AI named in your answers that you do not track yet. See [Competitors](/docs/setup/competitors). > [accordion] How do I add or change prompts? > > Open **AI Visibility** > **Prompts** and click **Add prompts**. Pick **Generate with AI**, **Manual/CSV** (type prompts or upload a CSV file) or **Google Search Console**. To stop tracking a prompt, archive it. See [Prompts and topics](/docs/setup/prompts-and-topics). > [accordion] What happens when I reach my prompt limit? > > Your plan tracks up to its daily prompt limit and no further, so your bill does not change by surprise. To track more, use **Add more prompts** in **Settings** > **Billing & plan**, or move to a bigger plan. If you have more Active prompts than your plan allows, for example after a downgrade, tracking pauses until you archive the extra prompts. > [accordion] Is there a free trial? > > Yes. The Basic plan has a 7-day free trial if your account can get one. You enter a card at checkout, and on day 7 the trial turns into a paid Basic plan by itself. Cancel before then if you do not want to continue. See [Plans and billing](/docs/account/plans-and-billing). > [accordion] How do I cancel? > > Go to **Settings** > **Billing & plan** and click **Manage subscription** on your current plan. Cancel in the Stripe page that opens. You keep access until the end of the billing period. See [Cancel your plan](/docs/account/plans-and-billing#cancel-your-plan). > [accordion] Where do I find my invoices? > > In **Settings** > **Billing & plan**, under **Billing**, click **Invoices**. Only a Company Admin can open billing. > [accordion] Can I track more than one brand? > > Yes, if your plan allows it. Basic includes 1 project, Standard 2 and Advanced 5. Add one from the brand switcher with **Add new brand**. See [Brands](/docs/setup/brands). > > Agency plans include 5, 10 or 25 client projects plus 3, 5 or 7 active pitch projects (Agency Basic, Agency Plus, Agency Max). See [Agencies](/docs/setup/agencies). > [accordion] Can I invite my team? > > Yes. Every plan includes unlimited team seats. Invite people from **Settings** > **Members** and give them a role: **Company Admin**, **Member** or **Viewer**. See [Team](/docs/setup/team). > [accordion] How is my data handled, and how do I delete it? > > See the [Privacy Policy](https://webglazer.com/privacy). To delete your account or company data, email [support@webglazer.com](mailto:support@webglazer.com). To remove a single brand, use **Delete project** in **Settings** > **Projects**. ## Still stuck? Email [support@webglazer.com](mailto:support@webglazer.com). # WebGlazer docs Source: https://webglazer.com/docs/index What WebGlazer is, which AI engines it tracks, and where to start in these docs. # Documentation WebGlazer shows you how AI assistants talk about your brand. > [card] [Quickstart](/docs/quickstart) > > Go from sign up to your first results. > [card] [Core concepts](/docs/concepts) > > Brands, prompts, topics, engines, runs and credits in plain words. > [card] [Metrics](/docs/metrics/visibility) > > Visibility, Share of Voice, mention order, sentiment and citations. > [card] [Dashboard overview](/docs/dashboard/overview) > > Read the Overview, Prompts, Platforms and the other pages. > [card] [Setup](/docs/setup/prompts-and-topics) > > Prompts, topics, competitors, engines, brands and your team. > [card] [Connect your AI agent](/docs/integrations/mcp) > > Connect your own AI agent to your WebGlazer data. ## What WebGlazer tracks WebGlazer shows you how AI assistants talk about your brand. Every day it asks ChatGPT, Perplexity, Google AI Mode, Gemini and Copilot the questions your buyers ask, then shows where you are named, which competitors win instead, and which sources the answers cite. WebGlazer collects answers from five AI engines: * ChatGPT * Perplexity * Google AI Mode * Gemini * Copilot WebGlazer asks each engine your prompts every day and records the answers. Results update once a day. The Basic plan tracks any 3 of these engines. Standard and Advanced track all 5. Agency plans choose 1 to 5. See [Plans and billing](/docs/account/plans-and-billing). For each answer, WebGlazer records whether your brand is named, where it appears in the answer, the tone used about it, which competitors are named, and which web pages the engine cites. ## Use these docs with AI Every page in these docs is also available in formats an AI assistant can read: * **[llms.txt](https://webglazer.com/docs/llms.txt)**: an index of every docs page with a one-line description. * **[llms-full.txt](https://webglazer.com/docs/llms-full.txt)**: the full text of every page in one file. * **Markdown pages**: add `.md` to the end of any page URL to get it as plain Markdown, for example `https://webglazer.com/docs/quickstart.md`. * **Docs MCP server**: connect an AI client to `https://webglazer.com/docs/mcp` so it can search these docs. The docs MCP server only searches these docs. To let an AI agent read your own brand data, use the [WebGlazer MCP server](/docs/integrations/mcp). ## Next steps > [card] [Quickstart](/docs/quickstart) > > Go from sign up to your first results. > [card] [Core concepts](/docs/concepts) > > Brands, prompts, topics, engines, runs and credits in plain words. > [card] [Dashboard overview](/docs/dashboard/overview) > > Read the Overview, Prompts, Platforms and the other pages. > [card] [Connect your AI agent](/docs/integrations/mcp) > > Connect your own AI agent to your WebGlazer data. # Connect your AI agent (MCP) Source: https://webglazer.com/docs/integrations/mcp Connect Claude, Claude Code, Cursor or another MCP client to your WebGlazer data, and see every tool it can use. The WebGlazer MCP server lets your own AI agent read your WebGlazer AI-visibility data: runs, prompts, AI answers, citations, sentiment, trends, products and your To-do list. You ask questions in plain language, and the agent answers from the same data the dashboard shows. One connection covers every brand you have access to. The agent can read everything listed below. It can change one thing: the status of items on your To-do list. ## Connection details | Property | Value | | - | - | | **Server URL** | `https://webglazer-mcp-56836240603.europe-west1.run.app/mcp` | | **Transport** | Streamable HTTP | | **Authentication** | OAuth sign-in, or a token you generate in WebGlazer | Always copy the current server URL from the dashboard: go to **Settings** > **Integrations** > **Your AI agent**, expand **Setup guide** and copy it from **Server URL**. If the URL in the dashboard differs from the one on this page, use the dashboard's. ## Before you start * A WebGlazer account with at least one brand. * An active subscription on the brands you want the agent to read. A brand whose subscription has ended is refused. * An AI client that supports MCP. In Claude, custom connectors work on every plan, but a Free plan can add only one custom connector. ## Two ways to connect **Sign in from your client (OAuth).** This is the recommended way. You paste the server URL into your client, your browser opens a WebGlazer approval screen, and you click **Allow**. There is no token to copy or keep safe. Use this for Claude.ai, Claude Desktop and Claude Code. **Paste a token.** For clients that cannot sign in, such as Cursor. You generate a token in WebGlazer and add it to your client as an `Authorization: Bearer` header. Both kinds of connection can read every brand you have access to and update the status of items on your To-do list. ### What you see on the approval screen When your client sends you to WebGlazer, you land on `app.webglazer.com/oauth/consent`. 1. If you are not signed in, the screen says **Connect your AI client to WebGlazer**. Click **Continue with Google**, or **Use email instead**. 2. The screen then asks **Connect \[client name]?** If WebGlazer could confirm where the client comes from, it shows the client as verified with its web address. If it could not, it warns that the name is unverified, because anyone who registers a client can pick any name. 3. It shows where your tokens will be sent (the client's web address). 4. It lists what the client is asking for: **Read your visibility data, prompts, citations and sentiment** and **Update the status of items on your To Do list** (the screen's wording for the To-do list). 5. It lists your brands under **This connection can read every brand you have access to:**. You do not pick a brand. Approving connects your whole account, and the connection covers every brand on that list. 6. Click **Allow** to connect, or **Deny** to cancel. If you have no brands yet, the screen says so and **Allow** is disabled. Create a brand in WebGlazer first. > [!NOTE] > > The connection follows your access. If you gain access to a new brand later, the agent can read it too. If you lose access to a brand, the agent loses it too. ## Generate a token You only need a token for clients that cannot sign in. **1. Open Your AI agent** In the dashboard, go to **Settings**, then **Integrations**, and find **Your AI agent**. ![Your AI agent in Settings, Integrations, with Generate token, the Setup guide and Clients and tokens](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-settings-ai-agent.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=163205b55408add6b6349b6814cb2b65) Illustration with sample data **2. Generate the token** Click **Generate token**. Under **Name this token**, type a name of up to 40 characters, such as the client you will use it in. Click **Generate**. **3. Copy it now** The dialog shows **Your token is ready** with the **Token** and the **Server URL**. Copy the token right away. It is shown once and never stored, so you cannot see it again. **4. Add it to your client** Expand **Setup guide**. Under **Option 2: Paste a token**, the configs for Claude Code, Cursor and other clients already contain the token you just generated. Copy the one you need. A token expires 30 days after you create it. Generate a new one when it does. > [!WARNING] > > Treat a token like a password. Anyone who has it can read every brand you have access to and change your To-do list. If a token leaks, revoke it. ## Connect your client > [tab] Claude.ai and Claude Desktop > > Copy the server URL from **Settings** > **Integrations** > **Your AI agent**: expand **Setup guide** and copy it from **Server URL**. > > Claude signs in with OAuth, so you do not need a token. > > **1. Add a custom connector** > > In Claude, go to **Customize** > **Connectors**. Click **+ Add**, then **Add custom connector**. > > **2. Enter the server details** > > Give the connector a name, such as WebGlazer. Paste the server URL you copied and click **Continue**. > > **3. Choose how to sign in** > > Under **Authentication**, choose **Sign in now** or **Sign in when needed**. Do not choose **No sign in**. Under **OAuth client**, keep **Use Claude's published identity (Recommended)**. Click **Add**. > > **4. Approve in WebGlazer** > > Your browser opens the WebGlazer approval screen. Check the brand list and click **Allow**. > > **5. Turn it on in a chat** > > In a conversation, click **+** in the lower left, open **Connectors** and switch WebGlazer on. > > On Team and Enterprise plans, an Owner or Primary Owner adds the connector first under **Organization settings** > **Connectors**. Members then go to **Customize** > **Connectors**, find the connector and click **Connect**. > > Source: Anthropic's guide, [Get started with custom connectors using remote MCP](https://support.claude.com/en/articles/11175166-get-started-with-custom-connectors-using-remote-mcp). > [tab] Claude Code > > Copy the server URL from **Settings** > **Integrations** > **Your AI agent**: expand **Setup guide** and copy it from **Server URL**. The examples below show today's URL; use yours if it differs. > > **Sign in (recommended).** Run this in your terminal, with your server URL: > > ```bash theme={null} > claude mcp add --transport http webglazer https://webglazer-mcp-56836240603.europe-west1.run.app/mcp > ``` > > Then start Claude Code, type `/mcp`, select the webglazer server and choose **Authenticate**. Your browser opens the WebGlazer approval screen. Click **Allow**. No token and no header are needed. > > **With a token.** Generate a token first, then run this, with your server URL: > > ```bash theme={null} > claude mcp add --transport http webglazer https://webglazer-mcp-56836240603.europe-west1.run.app/mcp --header "Authorization: Bearer " > ``` > > Source: [Connect Claude Code to tools via MCP](https://code.claude.com/docs/en/mcp). > [tab] Cursor > > Copy the server URL from **Settings** > **Integrations** > **Your AI agent**: expand **Setup guide** and copy it from **Server URL**. The examples below show today's URL; use yours if it differs. > > Connect Cursor with a token. > > **1. Generate a token** > > Follow [Generate a token](#generate-a-token) above. > > **2. Add the server to mcp.json** > > Add this to `.cursor/mcp.json` in your project (for one project) or to `~/.cursor/mcp.json` (for every project). Use your server URL, and replace `` with your token. > > ```json theme={null} > { > "mcpServers": { > "webglazer": { > "url": "https://webglazer-mcp-56836240603.europe-west1.run.app/mcp", > "headers": { > "Authorization": "Bearer " > } > } > } > } > ``` > > **3. Check the connection** > > Ask Cursor's agent to list your WebGlazer brands. > > Source: [Cursor MCP docs](https://cursor.com/docs/mcp). > [tab] Other clients > > Copy the server URL from **Settings** > **Integrations** > **Your AI agent**: expand **Setup guide** and copy it from **Server URL**. The examples below show today's URL; use yours if it differs. > > WebGlazer has not tested ChatGPT, VS Code, Windsurf or Codex yet. They may work through one of the ways below. > > **Clients that support MCP connectors with sign-in.** Add the server URL as a custom connector and sign in when your client asks. You land on the same WebGlazer approval screen. > > **Clients that support remote servers with headers.** Point the client at the server URL and send your token as the header `Authorization: Bearer `. > > **Clients that only launch a command.** Use `mcp-remote` to bridge to the server URL. Use your server URL, and replace `` with your token: > > ```json theme={null} > { > "mcpServers": { > "webglazer": { > "command": "npx", > "args": [ > "-y", > "mcp-remote", > "https://webglazer-mcp-56836240603.europe-west1.run.app/mcp", > "--header", > "Authorization: Bearer " > ] > } > } > } > ``` ## Check that it works Ask your agent: > *"List my WebGlazer brands."* The agent calls `list_projects` and shows each brand with its name, domain and your role. Then ask: > *"Show me the latest runs for \[brand]."* If both come back, you are connected. ## Example prompts * *"How did our AI visibility change week by week over the last three months?"* * *"Which sources does AI cite most when it answers our prompts? Are any of them sites where we are not mentioned?"* * *"How does AI describe our brand? Summarize the sentiment for the latest run."* * *"Show me what each engine answered for the prompt 'best project management tool for agencies'."* * *"Which prompts have the lowest visibility in the last run?"* * *"What is on our To-do list? Pick the three items that affect the most prompts and mark them as todo."* * *"Compare our visibility in Google AI Mode and ChatGPT for the last 30 days."* * *"Which of our products are most visible in ChatGPT?"* * *"Which competitor products show up in the same answers as our best seller?"* * *"Compare visibility across all my brands for the latest run."* ## Tools The server has 15 tools. All of them read data except `mark_opportunity`, which changes the status of an item on your To-do list. Every tool except `list_projects` takes a `project_id`: the brand to read, as `list_projects` returns it. Your agent handles this for you; you only need to name the brand in your question. A run is one day's set of AI answers for a brand. Some tools read one run, picked from `list_runs`. See [Core concepts](/docs/concepts#run). | Tool | What it returns | Parameters | Access | | - | - | - | - | | `list_projects` | The brands this connection can read: id, name, main domain and your role (admin, member or viewer). | None | Read | | `list_runs` | The brand's runs that have results, newest first, and the run the dashboard opens. | `project_id` | Read | | `get_run_summary` | Summary numbers for one run (mentions, answers, status), or for a date range. | `project_id`, the run, period and filters | Read | | `get_prompts` | The prompts in a run, the same list the dashboard shows, with each prompt's visibility numbers. | `project_id`, the run, `limit` (default 50, up to 100), `offset` | Read | | `get_prompt_answers` | The AI answers recorded for one prompt across engines. | `project_id`, the run, `prompt_id`, period and filters | Read | | `get_citations` | The citation leaderboard: the sources AI cited, plus the citation totals for the period. | `project_id`, `limit` (default 25, up to 100), `offset`, period and filters, `today` | Read | | `get_sources` | The same rows and totals as `get_citations`, listed as sources by host. | Same as `get_citations` | Read | | `get_sentiment` | How AI describes the brand: the sentiment breakdown for a run or a date range. | `project_id`, the run, period and filters | Read | | `get_trend` | Visibility over time, per run, day, week or month, as the dashboard trend shows it. | `project_id`, `granularity` (`run`, `day`, `week` or `month`), period and filters | Read | | `get_filter_options` | The topics, intents and engines that exist in a period, to use as filters in the other tools. | `project_id`, `date_from`, `date_to`, `all_time` | Read | | `list_opportunities` | Your opportunities and To-do items: cited sources where your brand is absent, ranked by how many prompts they affect. Dismissed items are left out. | `project_id`, period and filters, `today` | Read | | `get_opportunity` | One item with its evidence and a ready-to-run instruction for fixing it. | `project_id`, `id`, period and filters, `today` | Read | | `mark_opportunity` | Sets the status of one item and returns the new status. | `project_id`, `id`, `status` (`open`, `todo`, `done` or `dismissed`) | **Write** | | `list_products` | Product visibility for your connected Shopify store: each product's visibility, appearances, position, wins and win rate, merchants and prices, plus product categories. | `project_id`, `date_from`, `date_to`, `all_time`, `today`, `models`, `scope`, `sort`, `order`, `category`, `page`, `page_size` (up to 100) | Read | | `get_product` | One product: its numbers, visibility per linked prompt, the merchants AI named, competing products and its position distribution. | `project_id`, `shopify_product_id`, `date_from`, `date_to`, `all_time`, `today`, `models`, `scope` | Read | ### Period and filters Many tools accept the same optional period and filters: * `date_from` and `date_to`: the first and last day, as YYYY-MM-DD. Send both or neither. * `all_time`: `true` reads the dashboard's All time period. It cannot be combined with dates. * `topics`, `intents` and `models`: keep only these topics, intents or engines. Get the valid values from `get_filter_options`. * `today`: your local date, as YYYY-MM-DD. With no dates, the tool reads the period the dashboard opens on, and `today` makes that period match the one you see in the dashboard. Every result repeats the period it applied, so your agent can tell you which dates the numbers cover. The product tools take `models` but not `topics` or `intents`. Their `scope` is `shopping` (the default: only answers that showed product cards) or `all` (every answer). Product tools work only on plans that include Products: Standard, Advanced, agency plans, and the Basic free trial. On a paid Basic plan they return "Product visibility is not included in this project's WebGlazer plan, so there is no product data to show." With no store connected, `list_products` returns an empty list and `get_product` says the project has no Shopify store connected. Connecting or disconnecting a store happens in the dashboard. See [Products](/docs/dashboard/products) and [Plans and billing](/docs/account/plans-and-billing). ## Read and write access Every tool reads data except `mark_opportunity`. It sets an item to `todo` (picked to work on), `done` or `dismissed`, or reopens it with `open`. The change shows on your own To-do list in the dashboard. To-do statuses are personal, so teammates keep their own. It only writes to the brand you name. Every role, Viewers included, can change their own statuses. A token you generate in WebGlazer always includes write access. A client that signs in gets the permissions you see on the approval screen. If the screen did not include **Update the status of items on your To Do list**, the connection can still read everything, but `mark_opportunity` is refused. ## Limits * **One call, 45 seconds.** A tool call that takes longer stops with "WebGlazer took too long to answer; try again". * **Paged lists.** `get_prompts` returns 50 rows by default, `get_citations` and `get_sources` return 25. Each returns up to 100 rows per call. The agent pages through the rest with `offset`. * **Products.** `list_products` returns up to 100 products per page. * **Sign-in sessions.** A client that signed in renews its access on its own, but only for 90 days from the day you clicked **Allow**. This holds even if you use it every day. After 90 days, connect it again. * **Tokens.** A generated token expires after 30 days. ## Revoke access **1. Open the list** Go to **Settings** > **Integrations** > **Your AI agent** and expand **Clients and tokens**. It shows how many are active, for example "2 active of 3". **2. Find the connection** Under **All brands you have access to**, each row shows its name, a **Signed in** badge (a client that connected with OAuth) or a **Token** badge (a token you generated), when it was created and when it expires. **3. Revoke it** Click **Revoke**, then **Revoke?** to confirm. You see **Access revoked**. Access ends at once. The next call from that client fails, and a signed-in client cannot renew its access. You see and revoke only your own connections. If your WebGlazer account is deleted (you ask for this by emailing support), all of them are revoked. ## Troubleshooting > [accordion] My client says unauthorized (401) > > The token or sign-in is missing, expired or revoked. Check **Clients and tokens** under **Your AI agent**. If the row says **Revoked** or **Expired**, generate a new token or connect the client again. With a token, check that your client sends the header exactly as `Authorization: Bearer `. > [accordion] "project\_id is required: call list\_projects first" > > The agent called a tool without naming a brand. Ask it to list your WebGlazer brands first, then repeat your question with the brand name. > [accordion] "no brand with that id on this connection" > > You do not have access to that brand, or it no longer exists. Ask the agent to list your brands and use one from that list. > [accordion] "this connection covers one brand; reconnect to use all brands" > > This is an older connection made for a single brand. Connect the client again, or generate a new token, to get a connection that covers every brand. Then revoke the old one: it is listed under **One brand each** in **Clients and tokens**. > [accordion] "The WebGlazer subscription for this project has ended" > > The brand's subscription is no longer active. The project owner can renew it in WebGlazer under **Settings** > **Billing & plan**. Other brands on the same connection keep working. See [Plans and billing](/docs/account/plans-and-billing). > [accordion] "WebGlazer could not check this project's subscription. Try again shortly." > > WebGlazer could not check the brand's subscription at that moment. This is on WebGlazer's side. Wait a minute and ask again. If it keeps happening, email [support@webglazer.com](mailto:support@webglazer.com). > [accordion] "Product visibility is not included in this project's WebGlazer plan" > > The brand's plan does not include Products. Products is part of Standard, Advanced and agency plans, and the Basic free trial. See [Plans and billing](/docs/account/plans-and-billing). > [accordion] "This company is not active right now. Try again later." > > The company is temporarily locked, for example during an ownership transfer or a pending deletion. Try again later. > [accordion] "this connection is not authorized for mcp:write" > > Your client signed in with read access only, so it cannot change To-do statuses. Connect it again and make sure the approval screen lists **Update the status of items on your To Do list** before you click **Allow**. > [accordion] "WebGlazer took too long to answer; try again" > > The call ran longer than 45 seconds. Ask again, or narrow the question, for example to a shorter date range. > [accordion] The approval screen says "Invalid connection link" > > The link is incomplete or expired. Start the connection again from your AI client. > [accordion] The results are from an older run > > When the newest run has no data yet, the tools return the last run that does. The agent is told to say so and to name the date of that run. ## This server vs the docs MCP WebGlazer has two MCP servers. They do different things: * **The product MCP server** (this page) reads your own WebGlazer data. It needs you to sign in or use a token. * **The docs MCP server** at `https://webglazer.com/docs/mcp` searches these help docs only. It has no access to your account or data. Add it if you want your agent to answer "how do I" questions about WebGlazer. # Citations Source: https://webglazer.com/docs/metrics/citations Which websites AI engines cite when they answer your prompts, and how WebGlazer counts Times linked, Reach and the Mentioned? column. A citation is a web page an AI answer actually used. Citations show you which sites shape the answers about your market, so you know where to get featured. ## What it measures WebGlazer groups citations by source, meaning one website host. A subdomain is its own source, so `blog.example.com` and `example.com` are two rows. For each source you get how often AI linked to it, which engines did, how many of your topics it reaches, and whether its pages mention you. Search results an engine looked at but did not use are stored but never counted. ## How it is calculated **Times linked** = the number of answers that cited the source. An answer that cites three pages of the same site counts once for that site. > [!NOTE] > > **Example with made-up numbers.** ChatGPT cites 2 pages of `example.com` in one answer, and Perplexity cites 1 page of it in another answer. Times linked = 2. **Reach** = topics in which the source was cited ÷ all your topics in the view × 100, as a whole percent. The column draws it as five bars, one per 20%. Hover for the exact figure, for example "Cited across 3 of 10 topics · 30%". **Mentioned?** shows how many of the source's checked pages name you, as "found / checked", for example "2 / 5". WebGlazer reads up to the 5 busiest checked pages of each source. The pill's color and icon tell you the result at a glance: | What you see | Meaning | | - | - | | Green pill with a check mark | Every checked page names you. | | Amber pill with a check mark | Some checked pages name you. | | Red pill with an X | No checked page names you. | | Gray **Unknown** | No page on this source has been checked yet. | Hover the pill for the full sentence. ### How pages get checked WebGlazer checks pages of qualifying sources for your brand on its own. Some pages cannot be read this way. To fill those in, open a source on the **Citations** tab, switch to the full page list and click **Check more pages**. See [Citations tab](/docs/dashboard/citations#source-detail). Your own site shows **Your site** instead, because checking whether your own page names you means nothing. **Qualifying** sources are the ones cited by at least 3 different engines across at least 3 different prompts, counting non-branded prompts only. A source cited once by one engine is too thin to act on, so it stays off the default list. Qualifying sources get a rank (the # column), most linked first. **Category** is the kind of site, set per source from the newest run: **Your site**, **News & blogs**, **Community & social**, **Competitor site**, **Comparison sites**, **Official sources**, **Paid** or **Other**. A competitor site is a site on the domain of a competitor you declared. Which answers count: successful answers from the run's selected engines, over every run in the selected date range, narrowed by the topic, tag and platform filters. **Times linked** and **Reach** include answers to branded prompts. Only the qualifying bar counts non-branded prompts alone. ## Where you see it ![Citations tab source table with Cited by, Times linked, Reach and Mentioned?](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/wg-citations-2026-11.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=1debe37c8f082825e8738980f4897c9f) Sample data based on a linear.app project * **AI Visibility** > **Citations** tab: the source table, the category bar and the source detail sheet. See [Citations tab](/docs/dashboard/citations). * **AI Visibility** > **Platforms** tab: the **Citations** column, times linked per engine. * **AI Visibility** > **Opportunities** tab: qualifying sources that cite your market but do not name you. * The prompt detail sheet > **Sources** tab: the pages cited for one prompt. ## How to read it * Start with qualifying sources whose **Mentioned?** pill is red or amber. AI keeps going back to them and they do not talk about you yet. * High **Reach** means the source matters across many of your topics, not only one. * Check **Cited by**. A source cited by every engine is worth more than one cited only by one. ## What moves it * Getting your brand onto the pages AI cites: reviews, comparison lists, directories and community threads. * Publishing pages that answer the searches AI runs, so your own site gets cited. See [Query Fanouts](/docs/dashboard/fan-out). * Your prompts and topics. They decide which answers, and so which sources, are in the pool. # Mention order Source: https://webglazer.com/docs/metrics/mention-order Where AI first names your brand among the brands in an answer, averaged over your answers. Mention order is the average place where AI first names your brand among all the brands in an answer. 1 means AI named you first. Lower is better. ## What it measures Being named is good. Being named first is better, because readers often stop at the first option. Mention order tells you how early you show up in the answers that do mention you. In some places the same number is labeled **Position** (the stats line above the **Prompts** table). It is the same metric. ## How it is calculated For each answer, an AI model reads the answer and gives each brand it names a place: 1 for the first brand named, 2 for the second, and so on. Your brand's place in that answer is its place in this list. Every brand in the answer takes a place, including brands you do not track. So you can read 2.0 while the brand ahead of you is not in your **Rankings**. Mention order = sum of your places ÷ number of answers that gave you a place, rounded half up to one decimal (for example 1.8). * Only answers that name you count. An answer that does not name you adds nothing, so Mention order never punishes you for being absent. Read it next to [Visibility](/docs/metrics/visibility). * Only successful answers from the run's selected engines, to non-branded prompts. * Every run in the selected date range is pooled, then divided once. * If you merged a brand into your own, the earlier of the two places counts for that answer. > [!NOTE] > > **Example with made-up numbers.** Your brand is named in 4 answers. It comes 1st, 3rd, 1st and 2nd. Mention order = (1 + 3 + 1 + 2) ÷ 4 = 1.75, shown as 1.8. Competitors get a Mention order by the same rule. ## Where you see it ![Rankings table with the Mention order column for each brand](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-rankings.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=3c3667426ee4999b5d95971ef914aa3d) Illustration with sample data * **Overview** > **Rankings**: the **Mention order** column. * **AI Visibility** > **Visibility** tab: the **Mention order** line in the **Visibility Score** chart. The chart flips this line so up always means better. * **AI Visibility** > **Prompts** tab: the **Position** figure above the table and the **Mention order** column per prompt. * **AI Visibility** > **Platforms** tab: the **Mention order** column, one row per engine. * The prompt detail sheet: the **Mention order** card and the **Mention order** column of **Top Brands**. ## How to read it * 1.0 means AI named you first every time it named you. * A dash means no answer in the view gave you a place. * The change chip compares with the previous period of the same length. A lower number is an improvement, so a negative change shows as good. * Mention order is not the same as **Rank** (the # column in **Rankings**). Rank orders brands against each other, mostly by visibility. Mention order is where you appear inside each answer. ## What moves it * Being the brand that sources describe first, for example the top pick on comparison and review pages AI cites. * Clear category pages that answer the prompt directly, so AI reaches for your brand first. * Check the **Mention order** column on the **Platforms** tab to see which engine names you latest. # Sentiment Source: https://webglazer.com/docs/metrics/sentiment How positively AI answers talk about your brand, on a 0 to 100 score, and how WebGlazer calculates it. Sentiment is a 0 to 100 score for how well AI answers speak about your brand. 0 is all bad, 100 is all good. It tells you whether being named is helping or hurting you. ## What it measures Two answers can both name you, one as the top pick and one as "expensive and hard to set up". Sentiment captures that difference. WebGlazer scores your own brand only. Competitors do not get a sentiment score. ## How it is calculated 1. For each answer that names you, an AI model reads the answer, picks out the quotes about your brand and scores each one from 0 to 100, in steps of 10. 2. Each answer becomes one score: the average of its quote scores, snapped to the nearest step of 10. 3. Sentiment = the average of those per-answer scores, rounded half up to a whole number. Each answer counts once, however many quotes it has. Which answers count: * Successful answers from the run's selected engines. * Answers to **branded** prompts count here, unlike Visibility. A question that names you still shows how AI describes you. * Every run in the selected date range is pooled. * Topic, tag and platform filters apply. The score falls into one band, which sets its color: | Score | Band | | - | - | | 70 to 100 | Positive | | 50 to 69 | Neutral | | 0 to 49 | Negative | > [!NOTE] > > **Example with made-up numbers.** Three answers name you. Answer 1 has two quotes scored 80 and 90, so it scores 90 (85 snapped up). Answer 2 has one quote scored 60. Answer 3 has one quote scored 70. Sentiment = (90 + 60 + 70) ÷ 3 = 73.3, shown as 73, which is Positive. ### Category scores on the Sentiment tab The **Sentiment** tab also scores each category (for example **Pricing & Value** or **Support & Service**). Each quote is tagged positive, neutral or negative. A category's score = (positive quotes + half of neutral quotes) ÷ all quotes in the category × 100, rounded to a whole number. > [!NOTE] > > **Example with made-up numbers.** **Pricing & Value** has 6 positive, 2 neutral and 2 negative quotes. Score = (6 + 1) ÷ 10 × 100 = 70. ## Where you see it ![Sentiment tab with the score ring and category tones](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/wg-sentiment-2026-11.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=a3d047de5e6f2c301ac25324cb610bb5) Sample data based on a linear.app project * **AI Visibility** > **Sentiment** tab: the score ring, the summary and the category table. See [Sentiment tab](/docs/dashboard/sentiment). * **Overview** > **Rankings**: the **Sentiment** column, on your own row only. * **AI Visibility** > **Visibility** tab: the **Sentiment** line in the **Visibility Score** chart. * **AI Visibility** > **Prompts** tab: the **Sentiment** figure above the table, with a dot in the band color. * **AI Visibility** > **Platforms** tab: the score per engine, with a bar showing the positive, neutral and negative mix. * The prompt detail sheet: the **Sentiment** card. ## How to read it * Read the band before the exact number. 68 and 71 are close, but they fall in different bands. * The change chip compares with the previous period of the same length, in points. On the **Sentiment** tab, a move of 3 points or more either way reads as warming up or cooling. * With fewer than 8 scored answers, the **Sentiment** tab says "Still early, from just a handful of answers." Treat the score as a first read. * A dash means no answer in the view named you with a score. ## What moves it * What the pages AI reads say about you: reviews, comparison pages, forums and news. * Fixing the complaint that comes up most. The **Sentiment** tab names it and links to the quotes behind it. * Branded prompts. They count here, so adding questions like "Is your brand worth it?" changes the pool. # Share of Voice Source: https://webglazer.com/docs/metrics/share-of-voice Your brand's share of the mentions of your brand and tracked competitors in AI answers, and how WebGlazer calculates it. Share of Voice (SoV) is your slice of the brand mentions in AI answers, counted across your brand and the competitors you track. Where [Visibility](/docs/metrics/visibility) asks "how often am I named?", Share of Voice asks "of the brands I compare myself with, how much of the mentions are mine?". ## What it measures Share of Voice shows how much of the conversation you own compared with the competitors you track. It is the number to watch in a crowded market, where many answers list several brands at once. ## How it is calculated Share of Voice = answers that name your brand ÷ total brand mentions × 100, rounded half up to one decimal. "Total brand mentions" counts each brand once per answer that names it, added up over your brand and every competitor you track. So an answer that names 3 brands adds 3 to the total. An answer that repeats one brand's name 5 times still adds 1 for that brand. The same answer rules apply as for Visibility: * Only successful answers from the run's selected engines. * Only answers to non-branded prompts. * Every run in the selected date range is pooled, then divided once. * Topic, tag and platform filters apply to both sides. * The total counts your brand and every competitor you track. Brands you have not added, and brands you hid, are not in it. See [Competitors](/docs/setup/competitors). * Merged brands count as one brand. > [!NOTE] > > **Example with made-up numbers.** 10 answers count. Your brand is named in 5 of them. Competitor A is named in 6 and competitor B in 4. Total brand mentions = 5 + 6 + 4 = 15. Your Share of Voice = 5 ÷ 15 × 100 = 33.3%. Your Visibility for the same answers is 5 ÷ 10 = 50%. Competitors get a Share of Voice by the same rule, so the shares of your brand and your tracked, unhidden competitors add up to about 100% (rounding aside). ## Where you see it ![Rankings table with the SOV column for each brand](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-rankings.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=3c3667426ee4999b5d95971ef914aa3d) Illustration with sample data * **Overview** > **Rankings**: the **SOV** column, with a change chip. * **AI Visibility** > **Visibility** tab: the **Share of Voice** line in the **Visibility Score** chart. It is on by default. * **AI Visibility** > **Prompts** tab: the **Share of voice** figure above the table. ## How to read it * Visibility can rise while Share of Voice falls. That happens when answers name more brands overall: you appear more often, but so does everyone else. * A high Share of Voice with low Visibility means the answers that do name you name few other brands. * The change chip compares with the previous period of the same length, in percentage points (pp). On **All time** there is no chip. ## What moves it * Getting named in more answers moves it up, like Visibility. * Competitors getting named more moves it down, even if your own count stays the same. * Tracking more competitors adds their mentions to the total, so your share goes down. * Hiding a brand removes it from the total, so your share goes up. A hide by a Viewer changes only that Viewer's view. When a Member or Company Admin removes a tracked competitor in **Settings** > **Brands**, it leaves the total for everyone on the project. * Filters change the pool of answers, so your share in one topic can differ a lot from your overall share. # Visibility Source: https://webglazer.com/docs/metrics/visibility How often AI answers to your tracked prompts name your brand, and how WebGlazer calculates that percentage. Visibility is the share of AI answers that name your brand. It is the headline number in WebGlazer and the first thing to check when you want to know whether AI engines recommend you. ## What it measures Visibility answers one question: when people ask AI the questions you track, how often does the answer mention you? A higher number means more answers include your brand. It says nothing about where in the answer you appear or how the answer talks about you. For that, see [Mention order](/docs/metrics/mention-order) and [Sentiment](/docs/metrics/sentiment). ## How it is calculated Visibility = answers that name your brand ÷ answers counted × 100, rounded half up to one decimal (for example 42.5%). Which answers count: * Only answers that came back successfully from one of the engines selected for that run. An answer that errored is left out of both sides of the division, so a failed answer never counts as "not mentioned". * Only answers to non-branded prompts. A branded prompt already names your brand in the question, so it is left out of visibility. Branded prompts still count for [Sentiment](/docs/metrics/sentiment). * Every run in the selected date range is pooled. WebGlazer adds up the counts across all runs in the period and divides once. It never averages per-run percentages. * Topic, tag and platform filters narrow both the top and the bottom of the division. An answer counts as naming you once, however many times it repeats your name. > [!NOTE] > > **Example with made-up numbers.** In the last 7 days, 3 runs asked your 10 non-branded prompts on 4 engines: 120 answers. 2 answers errored, so 118 answers count. Your brand is named in 47 of them. Visibility = 47 ÷ 118 × 100 = 39.8%. Some views round differently: * **Per prompt** (the ring in the **Prompts** table and the prompt detail sheet): a whole percent, for example 18%. * **Per engine** (the **Platforms** tab): each engine divides by its own answers only. An engine that answered 1 prompt and named you reads 100%. * **Per topic** (the **Topics** tab): the same rule over the answers to that topic's prompts. Competitors get a visibility score by the same rule: answers naming that competitor ÷ answers counted. Only competitors you track get a score. A brand AI names that you do not track is not shown. See [Competitors](/docs/setup/competitors). ## Where you see it ![Visibility Score panel showing the visibility number and its trend line](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/wg-visibility-score-2026-11.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=95955c5a9a962be30c90e3ba692f2f6c) Sample data based on a linear.app project * **Overview** > **Visibility Over Time** chart and the **Visibility** column of **Rankings**. * **AI Visibility** > **Visibility** tab: the big number in the **Visibility Score** panel, and the trend line. * **AI Visibility** > **Prompts** tab: the **Visibility** figure above the table and the ring next to each prompt. * **AI Visibility** > **Topics** tab: tile shading and the **Visibility** column. * **AI Visibility** > **Platforms** tab: the **Visibility** column, one row per engine. * The prompt detail sheet: **Visibility Over Time** and the **Visibility** column of **Top Brands**. ## How to read it * Compare against your competitors in **Rankings**, not against 100%. Most answers name only a few brands. * The change chip next to the number compares with the previous period of the same length (for example the 7 days before the last 7 days). It shows in percentage points (pp). On **All time** there is no comparison. If the previous period had no answers to compare, you see a "no comparison" note instead of a number. * A prompt that never names you reads 0%. That is a real result, not missing data. * Few answers make the number jumpy. A filter down to one topic and one engine can swing a lot between runs. ## What moves it * Being named on the pages AI engines cite. See [Citations](/docs/metrics/citations) and the **Opportunities** tab. * Having pages that answer the searches AI runs before it answers. See [Query Fanouts](/docs/dashboard/fan-out). * Your prompt set. Adding or archiving prompts changes which answers count. * The engines you track. Each engine has its own visibility, and the blended number depends on which engines are selected. * Hiding a competitor never changes how many answers count. A Viewer who removes a brand hides it from their own view only. When a Member or Company Admin removes a tracked competitor in **Settings** > **Brands**, it is removed for everyone on the project. Merging a brand into your own brand adds the answers that name it to your count. # Quickstart Source: https://webglazer.com/docs/quickstart Set up WebGlazer from sign up to your first AI visibility results in the dashboard. This guide walks you through the setup wizard at [app.webglazer.com](https://app.webglazer.com), from creating your account to seeing your first results on the Overview. You need your website address and a payment card for the plan or free trial. Your first run starts as soon as you finish setup. The engines usually answer within about four to six minutes, and your results appear on the **Overview** when the run finishes. After that, your data updates once a day without you doing anything. ## Before you start * Have your website address ready, for example `yourbusiness.com`. * Know the market (country) and language your buyers search in. * Have a few competitors in mind. WebGlazer suggests some, and you can add your own. ## Set up tracking The whole flow at a glance: | Stage | What happens | | - | - | | Create your account | Sign up with Google or with email and a password. | | Steps 1 to 4 | Project details, brand, topics and prompts. | | Step 5 | Choose a plan in Stripe. | | After checkout | Pick your engines and invite your team. | | First run | About 4 to 6 minutes, then your results are on the Overview. | **1. Create your account** Go to [app.webglazer.com](https://app.webglazer.com) and choose **Continue with Google**, or enter your email and a password and click **Create account**. **2. Choose who you are tracking** On **Who are you tracking?**, pick **My own brand** to track your own company, or **My clients** if you are an agency. This guide follows **My own brand**. Agencies first fill in **Your agency** and then add a first **Client project**. See [Agencies](/docs/setup/agencies). **3. Step 1: Project details** The **Project details** step opens a screen headed **Brand setup**. Fill in: * **Website URL**: your main website. * **Brand name**: the name AI answers should use for you. * **Market** and **Language**: where your buyers are and the language they search in. WebGlazer may suggest these from your website. Click **Use this** to accept the suggestion. Click **Next**. WebGlazer creates your project and starts reading your website. ![Setup wizard step 1, Brand setup, with Website URL, Brand name, Market and Language](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-onboarding-project-details.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=15df272f4ec32e7b4db9c0fffbcd5372) Illustration with sample data **4. Step 2: Brand understanding** WebGlazer reads your site and fills in your **Brand profile**. Check and edit each part: * **Description** and **Category** * **What you sell**: your products or services. * **Who you sell to**: your buyers. * **Who you compete with**: suggested competitors. Remove any that are wrong, or use **Add a competitor**. * **Question mix**: how much weight to give each buyer stage: **Learning**, **Comparing** and **Buying**. The **About you** switch adds questions that name your brand. Click **Next**. Your changes save as you go. **5. Step 3: Topics** On **Review topics**, pick the subjects you want to watch in AI answers. The counter next to **Select topics** shows how many you can keep. Edit a topic name, delete one, or add your own with **New topic**. Click **Next**. **6. Step 4: Prompts** On **Review prompts**, check the questions WebGlazer will ask the AI engines, grouped by topic. Edit or delete any prompt, or type your own in **Add a custom prompt**. The counter next to **Select prompts** shows how many you keep against your limit. Click **Looks good**. Not sure which prompts to keep? Keep the ones a buyer would really ask, and leave your brand name out of most of them. See [Prompts and topics](/docs/setup/prompts-and-topics#write-prompts-that-measure-something-useful). **7. Step 5: Review and choose a plan** You see a summary: **Setup complete**, how AI describes you today, example questions, and the competitors you are **Benchmarked against**. Click **See my plan**. Pick a plan. You go to Stripe to enter your card. If your account can get the free trial, the Basic plan shows **Start 7-day free trial**. See [Plans and billing](/docs/account/plans-and-billing). When checkout is done, Stripe sends you back to WebGlazer to finish the last steps below. **8. Pick your AI engines** Back in WebGlazer, **Welcome to WebGlazer** shows the AI engines. On Standard and Advanced it says "All 5 AI models are in your plan" and every engine is on. On Basic it says "Pick up to 3 of the 5 AI models". On an agency plan you pick 1 to 5, and a line shows how many daily credits your choice uses. Click **Continue**. You can change engines later. See [Engines and schedule](/docs/setup/engines-and-schedule). **9. Answer a few questions and invite your team** Answer the short questions about your role and goals, then **How did you hear about WebGlazer?**. On **AI visibility is a team effort**, enter teammates' emails, pick a role, and click **Send invite**, or click **Skip for now**. The roles are **Company Admin**, Member (shown here as **Company Member**) and **Viewer**. See [Team](/docs/setup/team). On **Setup complete. You're ready to go.**, click **Explore insights**. **10. Watch your first run** Your first run starts now. The **Run Progress** popup opens over the Overview with one ring per AI engine. Each ring fills as answers come in. You can close the popup; the run keeps going. **11. Read your results on the Overview** When the run finishes, the **Overview** shows **Visibility Over Time** and the **Rankings** table, which compares your brand with your competitors on visibility, sentiment and mention order. ![Rankings table on the Overview after the first run](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-rankings.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=3c3667426ee4999b5d95971ef914aa3d) Illustration with sample data See [Overview](/docs/dashboard/overview) for how to read each part. > [!NOTE] > > After the first run, WebGlazer asks the AI engines your prompts once a day for each brand and records the answers. New data shows up after each daily run. You do not need to do anything to keep tracking. ## What to do next 1. Go to **Settings** > **Brands** and accept the competitors you want to compare with from **Brand suggestions**. Only your own brand and the competitors you track count in **Rankings**, Share of Voice and the other competitor numbers. A brand AI names that you have not accepted is left out. See [Competitors](/docs/setup/competitors). 2. Open **AI Visibility** > **Prompts** to see each answer and where your brand appears. 3. Invite colleagues from **Settings** > **Members**. ## Next steps > [card] [Core concepts](/docs/concepts) > > Brands, prompts, topics, engines, runs and credits in plain words. > [card] [Write better prompts](/docs/setup/prompts-and-topics) > > Pick the prompts and topics that measure something useful. > [card] [Read the Overview](/docs/dashboard/overview) > > Visibility Over Time and Rankings, part by part. > [card] [Connect your AI agent](/docs/integrations/mcp) > > Let Claude, Cursor or another MCP client read your WebGlazer data. # Agency accounts Source: https://webglazer.com/docs/setup/agencies How agency accounts work in WebGlazer: client and pitch projects, slot limits per plan, daily credits, and converting a pitch to a client. Agency plans let you track many client brands from one account and run short pitch projects for prospects before they sign. This page covers what changes when your account is an agency account. ## Clients and pitches Every project on an agency account is one of two kinds: * **Client**: tracked every day for as long as you work with them. * **Pitch**: a short project for a prospect. It runs for 7 days with up to 60 prompts and uses your shared credits. When the prospect signs, you convert it to a client. When you add a project, WebGlazer first asks **Client or Pitch?** and shows how many slots of each kind you have left. ## Slots per plan | Plan | Clients | Active pitches | | - | - | - | | Agency Basic | 5 | 3 | | Agency Plus | 10 | 5 | | Agency Max | 25 | 7 | A pitch counts against your pitch slots only while it is active. A slot frees up when a pitch ends or becomes a client. Brand plans (Basic, Standard, Advanced) have no pitch projects. ## Add a client or a pitch * From the brand switcher at the top of the sidebar, click **Add new client**. * Or go to **Settings** > **Projects**, open the **Clients** or **Pitches** tab and click **Add client** or **Add pitch**. When every client slot is used, the switcher shows **Client limit reached** with the count, for example 5/5, and an **Upgrade** link, and the Add buttons are off with the reason. The switcher lists your projects under **Clients** and **Pitches**. A pitch shows a **Pitch** badge and a countdown such as "Ends in 4 days", "Ends tomorrow" or "Ends today". After the 7 days it shows **Pitch ended**. ## Daily credits Agency plans measure usage in credits instead of a prompt count. You buy credits per month, and WebGlazer turns them into a daily amount: **Daily credits = monthly credits ÷ 31, rounded down.** Dividing by 31 means a full month of daily runs always fits in your monthly credits. | Plan | Credits per month | Daily credits | | - | - | - | | Agency Basic | 14,400 | 464 | | Agency Basic, Expanded | 21,600 | 696 | | Agency Plus | 30,600 | 987 | | Agency Plus, Expanded | 43,200 | 1,393 | | Agency Max | 70,200 | 2,264 | | Agency Max, Expanded | 91,800 | 2,961 | If you buy more credits with **Add more credits** in **Settings** > **Billing & plan**, your daily credits are worked out from the new monthly total. How credits are used: * Each prompt uses one credit per AI engine every day. A project with 20 prompts on 3 engines uses 60 credits a day. * Every client and active pitch shares the same credits. Ended pitches use none. * **Settings** > **Projects** shows **Used / max daily credits** for the whole account. There is no per-project credit column; the **Prompts** column shows each project's prompts. * The **Credits** meter in the sidebar shows what is left for the month. It warns when you are running low. When credits run out, runs pause until your credits reset for the next month. When you pick engines during setup, WebGlazer shows how many daily credits the choice uses and how many are left. If the choice needs more than you have, **Continue** stays off. The same check applies in **Tracking setup** and when you add prompts. ## Choose engines on an agency plan Agency plans choose 1 to 5 AI engines per project. More engines give you more answers per prompt and use more credits. See [Engines and schedule](/docs/setup/engines-and-schedule). ## When a pitch ends After 7 days a pitch stops running. Its data stays readable. Convert it to a client to keep tracking. ## Convert a pitch to a client **1. Open the pitch** Go to **Settings** > **Projects** and open the **Pitches** tab. **2. Convert** Open the pitch's row menu and choose **Convert to client**, then confirm in **Convert to Client?**. Members and Company Admins can convert a pitch. Viewers cannot. The project becomes a client and uses one of your client slots. Daily tracking continues, an ended pitch starts tracking again, and its data stays. If all client slots are in use, upgrade your plan or delete a client first. # Multiple brands Source: https://webglazer.com/docs/setup/brands How to add another brand to your WebGlazer account, switch between brands, and what each brand keeps separate. One WebGlazer account can track several brands, for example a company with more than one product line or website. Each brand is tracked on its own, with its own prompts, competitors and results. In **Settings** a brand is called a project. ## How many brands your plan includes | Plan | Brands | | - | - | | Basic | 1 | | Standard | 2 | | Advanced | 5 | Agency plans count clients and pitch projects instead. See [Agency accounts](/docs/setup/agencies). ## Add a brand **1. Open the brand switcher** Click the brand name at the top of the sidebar. **2. Start the new brand** Click **Add new brand**. The item only shows while your plan has a free brand slot. Basic includes 1 brand, so on Basic the item never shows. To add a brand, move to Standard or Advanced. See [Plans and billing](/docs/account/plans-and-billing). **3. Fill in the first two steps** Enter the website and market under **Project details**, then confirm what WebGlazer found under **Brand understanding** (description, category, what you sell, who you sell to and competitors). If the website already belongs to a brand on your account, WebGlazer says so and offers to open it. **4. Create prompts** The dashboard switches to the new brand. It has no prompts yet, so a **Create prompts for** window asks you to add them. Click **Create prompts** to open Add Prompts, or **Later** to do it from **AI Visibility** > **Prompts**. A new brand is tracked daily once it has Active prompts. It copies the run time and time zone of your newest other brand, and its results appear after its daily run. See [Prompts and topics](/docs/setup/prompts-and-topics) and [Engines and schedule](/docs/setup/engines-and-schedule). A **Company Admin** who is not the account owner sees **New project for this company** in the switcher instead. ## Switch between brands Click the brand name at the top of the sidebar and pick another brand. Every page then shows that brand's data. The gear icon next to a brand opens its **Brand profile** settings. ## What each brand has Each brand keeps its own: * Website, **Market** and **Language**, and **Time zone** * Brand profile: **Description**, **Category**, **What you sell**, **Who you sell to** and **Question mix** * Topics and prompts * Tracked competitors and brand suggestions * AI engines * Results, reports and To-do list * Shopify store connection on [Products](/docs/dashboard/products) and [Search Console](/docs/dashboard/search-console) connection Shared across all brands on the account: * The prompt allowance. All brands draw from one pool, so prompts added to one brand leave fewer slots for the others. * Billing and plan. * Team members. Viewers and Members only see the brands they were given; a Company Admin sees all of them. See [Team](/docs/setup/team). ## Manage brands in Settings **Settings** > **Projects** lists every brand with its market, active prompts against your plan limit, AI engines, **Frequency** and creation date. Click a brand name to switch to it. The row menu has: * **Edit details**: market, language and time zone. * **Tracking setup**: AI engines. See [Engines and schedule](/docs/setup/engines-and-schedule). * **Export chats**: downloads the brand's AI answers as a CSV file. * **Delete project**: Members see the item, but only a **Company Admin** can delete a project. ## Delete a brand Open **Settings** > **Projects**, open the brand's row menu and choose **Delete project**. Type the brand's domain to confirm, then click **Delete this brand**. > [!WARNING] > > Deleting permanently removes the brand, its topics, prompts and run history. It cannot be undone. # Competitors Source: https://webglazer.com/docs/setup/competitors How WebGlazer finds competitor brands, how to add, accept, remove and restore them, and where they show up in the dashboard. Competitors give your numbers context. 15% visibility means more when you can see a rival at 45% on the same prompts. WebGlazer compares your brand with the competitors you track on every chart and ranking. ## Where you manage competitors Go to **Settings** > **Brands**. The page has two parts: * **Your brands**: a table of your own brand, marked **You**, and every competitor you track. Columns are **Brand**, **Domain** and **Mentions**. Mentions counts how often AI answers named the brand in the last 7 days. * **Brand suggestions**: a list on the right (below the table on small screens) of brands AI named in your answers that you do not track yet. The **Brands** item in the settings menu shows how many suggestions are waiting. You need the **Member** or **Company Admin** role to add competitors and to accept or reject suggestions. A **Viewer** sees **Brand suggestions** read-only. A Viewer can still use **Remove brand**, but it only hides the brand from that Viewer's own dashboard. Only your own brand and the competitors you track count in **Rankings**, Share of Voice, mention order lists and topic rankings. A brand AI names that you do not track is left out of all of them until you add it or accept it here. ## How competitors are found 1. **During brand setup.** On the **Brand understanding** step WebGlazer proposes competitors from your website. You can remove them or add your own before you continue. If no confident match is found, the step says so and you can add them by hand. 2. **From your answers.** After each run, WebGlazer collects every brand the AI engines named in answers to your prompts. Brands you do not track yet appear under **Brand suggestions**, most mentioned first, with their mention count. There is no limit on how many competitors a brand can track. ## Accept or reject a suggestion In **Brand suggestions**, each card has two buttons: * **Accept** (check mark): the brand joins your tracked list. It is tracked by name with no domain, so its **Domain** column stays empty. * **Reject** (cross): the brand is removed from suggestions and is not suggested again. The list shows 20 suggestions at a time. Click **Show more** to see the next ones. ## Add a competitor by hand **1. Open the dialog** In **Settings** > **Brands**, click **Add brand**. **2. Enter the brand** Fill in **Domain**, for example `acme.com`. **Brand name** is optional and lets you set the name to match; without it, WebGlazer uses a name based on the domain. **3. Save** Click **Add brand**. The brand is tracked from your next run, so it appears in the table and charts after that run finishes. ## Remove and restore a competitor Open the row menu of a competitor and choose **Remove brand**. What happens depends on your role: * **Member or Company Admin**: the competitor comes off the project's tracked list for everyone. You see "We stopped tracking ... and took it off your list." It no longer counts in anyone's numbers, even if a later run names it. * **Viewer**: the brand is hidden from your own dashboard only. You see "... was removed from your dashboard." Other people still see it. To bring a brand back, turn on **Show hidden** above the table. Removed, hidden and rejected brands appear there grayed out. Each row has one of two actions: * **Restore**: for a brand you hid yourself. It comes back on your dashboard. * **Track again**: for a brand that was taken off the tracked list or rejected from suggestions. It goes back on the list and counts again for everyone. Members and Company Admins can do this. Your own brand cannot be removed from this table. ## Where competitors show up * **Overview**: the **Rankings** table and the **Visibility Over Time** chart compare your brand with each tracked competitor. * **AI Visibility**: visibility, share of voice and mention order are shown for you and your tracked competitors. Sentiment is scored for your own brand only. * **To-do list and Opportunities**: each source gap lists the competitors that AI mentioned there instead of you, under "Mentioned here instead". * **Settings** > **Brands**: mention counts for the last 7 days. When there is no run in the last 7 days, the table cannot show mentions. You still see your tracked list with **Add brand** and **Remove**, so you can set up competitors before the first run. # Engines and schedule Source: https://webglazer.com/docs/setup/engines-and-schedule Which AI engines WebGlazer can track, how to choose them, how often runs happen, how to set the time zone, and how to follow a run. WebGlazer asks your prompts to a set of AI engines once a day and records every answer. This page explains which engines you can track, when runs happen and how to follow one. ## Engines you can track | Engine | Notes | | - | - | | ChatGPT | | | Perplexity | | | Google AI Mode | Can be slower than the others, so its results may arrive later or partly. | | Gemini | | | Copilot | | Most engines are asked from the country set in **Settings** > **Brand profile** > **Market and language**, when that engine supports the country. Google AI Mode also gets the language. ## How many engines your plan includes | Plan | Engines | | - | - | | Basic | Choose 3 of 5 | | Standard, Advanced | All 5, always on | | Agency Basic, Agency Plus, Agency Max | Choose 1 to 5 | On plans that include all 5, every run uses all of them and there is nothing to choose. On agency plans each engine you turn on uses daily credits, so the choice affects how many prompts you can run. See [Agency accounts](/docs/setup/agencies). ## Choose your engines ![Tracking Setup dialog with the prompts row, the five AI engines and the Summary](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-tracking-setup.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=66ea85592913a8ee38d8079d27a91d01) Illustration with sample data You choose engines when you set up a brand. On Standard and Advanced every engine always runs, so there is nothing to change. On Basic and agency plans, change them later like this: **1. Open Projects** Go to **Settings** > **Projects**. **2. Open Tracking setup** Click the engines in the brand's **AI engines** column, or open the row menu and choose **Tracking setup**. **3. Switch engines on or off** Under **AI engines**, switch each engine on or off. The counter shows how many you selected out of your plan's limit. The **Summary** on the right shows **Active prompts**, what is left in your plan, the engines and **Frequency**. **4. Save** Click **Update**. You need at least 1 engine. On an agency plan, **Update** stays off if the change needs more daily credits than you have left. You need the **Member** or **Company Admin** role to change engines. A change applies from the next run. ## How often runs happen Every brand runs once a day. In **Settings** > **Projects**, the **Frequency** column shows **Daily**. You do not start runs yourself: the daily run asks every Active prompt to every engine you track. The first run of a new account starts when you finish setup. After that, runs repeat every day at the same local time: * Your first brand runs every day at the time of day you created it, in your browser's time zone at that moment. * A brand you add later copies the run time and time zone of your newest other brand. * The dashboard does not show the time of day, and you cannot change it. You can change the time zone, as shown below. ## Set the time zone The time zone decides which calendar day a run belongs to and the local time it runs at. Each brand has its own. **1. Open Edit details** Go to **Settings** > **Projects**, open the brand's row menu and choose **Edit details**. **2. Pick a time zone** Choose a value under **Time zone**. The same dialog also holds the brand's **Market** and **Language**. **3. Save** Click **Update**. A "Project updated" message confirms it. Only a **Company Admin** can change the time zone. When you change it, the run keeps the same local time of day in the new zone. ## Follow a run When your first run starts at the end of setup, the **Run Progress** window opens over the Overview. It shows one ring per engine that fills as answers arrive, and the step each engine is on. If an engine fails or is delayed, a summary under the rings says which engine and on how many prompts. You can close the window at any time; the run keeps going. The window opens only for this first run. If the latest run is not finished or failed, the dashboard keeps showing your last completed results and a banner says so. # Prompts and topics Source: https://webglazer.com/docs/setup/prompts-and-topics How to add, edit, archive, restore and delete the prompts WebGlazer asks AI engines, and how to group them into topics. Prompts are the questions WebGlazer asks AI engines for your brand every day. Every number in the dashboard comes from the answers to these prompts, so the prompt list decides what you measure. ## How prompts and topics fit together * A **prompt** is one question, for example "Best project management tool for small teams?". WebGlazer sends it to each AI engine you track. * A **topic** groups related prompts, for example "Pricing comparisons". Topics let you filter the dashboard and see visibility per subject. * Each brand (project) has its own prompts and topics. You manage both on **AI Visibility** > **Prompts**. You need the **Member** or **Company Admin** role to change them. A **Viewer** sees the list but not the edit controls. ## Where your first prompts come from When you set up a brand, WebGlazer reads your website and suggests the rest: 1. **Topics**: WebGlazer generates topics for your brand. You tick the ones to keep, edit their names, or add your own. The counter shows how many topics your setup allows. 2. **Prompts**: WebGlazer writes prompts for each kept topic. You review them, edit or remove any, then click **Looks good**. After setup you add more prompts from the Prompts tab. ## Add prompts **1. Open the Prompts tab** Go to **AI Visibility** > **Prompts** and click **Add prompts**. **2. Pick a source** Choose one of three ways: * **Generate with AI**: WebGlazer suggests new prompts. You can pick **Topics (optional)** and fill in **What should we focus on?**, for example "pricing questions, comparisons with other tools". Click **Generate**. * **Manual/CSV**: choose **Type prompts** and enter one prompt per line, or choose **CSV file** and upload a file with a `prompt` column and an optional `topic` column. A file can have up to 1,000 rows. For example: ```csv theme={null} prompt,topic Which CRM works for a 10-person sales team?,CRM for small teams Best CRM with a free plan for startups?,Pricing comparisons ``` * **Google Search Console**: WebGlazer finds question-style searches people already make on Google for your site. A **Company Admin** connects Search Console first. Anyone else can upload a Search Console Performance export (CSV or ZIP, up to 20 MB) with **Upload GSC export instead**. **3. Review the list** Every source feeds one review list. Edit the text of any row, give each prompt a topic, or use **Select all**, **Assign topic** and **Remove from review** on several rows at once. Prompts you already track or already archived are skipped and counted at the top. **4. Add them** Click **Add selected**. The footer shows how many slots you have, for example "18 selected · 300 slots available → 282 after adding". If your selection is bigger than your free slots, **Add selected** stays off until you select fewer. New prompts are asked in the next daily run. Until then the Prompts table marks them **Not run yet**. > [!TIP] > > You can close the Add Prompts window while **Generate with AI** is working. The results wait for you when you open it again. ## Edit a prompt Open the row menu of a prompt in the **Active** tab and choose **Edit prompt**. You can change the text and the topic, then click **Save**. A prompt cannot be edited into the same text as another prompt you already track or archived. ## Archive and restore prompts Archiving stops a prompt from running but keeps its past results. * **Archive one prompt**: row menu > **Archive prompt**. * **Archive several**: tick the rows and click **Archive** in the bar that appears. * **Free up slots when you are at your limit**: click **Manage prompts**, tick the prompts to stop, then click **Archive N prompts**. Archived prompts move to the **Archived** tab of the Prompts table. Each row still shows its last result. Click **Restore** (or tick several and click **Restore selected**) to make a prompt Active again. Restoring needs a free slot. If the prompt's topic was archived, choose a new topic first. ## Delete prompts Deleting is only possible from the **Archived** tab. Click **Delete** on a row, or tick several and click **Delete selected**. > [!WARNING] > > Delete has no confirmation step and cannot be undone. It erases the prompt and all of its results, so your past numbers are recalculated without it. Adding the same text again later starts a new prompt with no history. If a run that asked the prompt is still saving answers, wait until it finishes. ## Manage topics On a wide screen the topic list sits left of the Prompts table. On a smaller screen click **Manage topics** to open it. * **All topics** shows every Active prompt. Click a topic to filter the table to it. * **Add topic**: enter a **Topic name** and click **Add topic**. If a topic with that name is archived, you are offered **Restore** instead. * **Rename**: click the edit icon next to a topic, change the name and click **Save**. The prompts stay in the topic. * **Remove**: click the remove icon. The prompts and their history stay. If the topic still has Active prompts, choose where they go under **Move prompts to** and click **Move prompts and remove**. ## How many prompts you can track Your plan sets how many Active prompts your account can track at once. All brands on the account share this one pool. Archived prompts do not count. | Plan | Active prompts included | | - | - | | Basic | 50 | | Standard | 100 | | Advanced | 200 | You can buy more prompts as an add-on in **Settings** > **Billing & plan**. Agency plans count daily credits instead of prompts: each prompt uses one credit per AI engine every day. See [Agency accounts](/docs/setup/agencies). With the add-on, Basic goes up to 200, Standard up to 250 and Advanced up to 350, in steps of 50. The Prompts table shows your current use as "used / allowance" next to **Add prompts**. At the limit, the button turns into **Upgrade plan** and **Manage prompts**. On an agency plan the count is in daily credits instead, for example "120 of 464 daily credits used", and you reach the limit when your daily credits are fully committed. ## Write prompts that measure something useful These tips follow from how WebGlazer asks your prompts and counts the answers. * **Ask like a buyer would.** AI engines answer the whole question, not keywords. "Which CRM works for a 10-person sales team?" tells you more than "CRM software". * **Leave your brand name out of most prompts.** A prompt that names your brand is marked as branded. It shows how AI describes you, but it cannot show whether AI recommends you on its own. The [To-do list](/docs/dashboard/todo) counts only prompts that do not name your brand. * **Cover the whole buying journey.** In **Settings** > **Brand profile**, the **Question mix** sets how often new AI-generated prompts are about **Learning**, **Comparing** or **Buying**. It applies to prompts you add from then on. Prompts you already track stay as they are. * **Check your market and language.** Most engines are asked from the country set under **Settings** > **Brand profile** > **Market and language**, and Google AI Mode also gets the language. Write prompts in the language your buyers use there. * **One idea per prompt.** Two prompts that ask the same thing in different words give you near-identical answers and use two slots. * **Use topics as your reporting groups.** Put prompts in the topic you want to read results for, since the dashboard filters by topic. * **Keep personal data out.** Prompts are sent to third-party AI services, so never include names, emails or other personal or sensitive data. # Team Source: https://webglazer.com/docs/setup/team How to invite teammates to WebGlazer, what each role can do, how to accept an invite, and how to change or remove access. Invite teammates so they can read or work on your brands. Each person gets one role in your company, and Viewers and Members only reach the brands you pick for them. ## Roles | What they can do | **Viewer** | **Member** | **Company Admin** | | - | - | - | - | | Projects they reach | The selected projects | The selected projects | Every company project | | Read results | Yes | Yes | Yes | | Keep their own [To-do list](/docs/dashboard/todo) | Yes | Yes | Yes | | Connect an AI agent through the [MCP server](/docs/integrations/mcp) | Yes | Yes | Yes | | Edit content, prompts, topics, competitors and tracking | No | Yes | Yes | | Invite and manage teammates | No | No | Yes | | Change billing and the plan | No | No | Yes | | Change a brand's time zone | No | No | Yes | | Connect Google Search Console or a Shopify store | No | No | Yes | | Delete projects | No | No | Yes | The **Owner** is the person who created the company. The owner holds the subscription and has full access. An AI agent connected through the MCP server sees only the brands that person can see. ## Invite a teammate **1. Open Members** Go to **Settings** > **Members** and click **Invite**. **2. Pick a role** Choose **Viewer**, **Member** or **Company Admin**. For a Viewer or Member, tick the **Projects** they should reach. A Company Admin reaches every project, so there is nothing to tick. **3. Send it** Enter their **Email** and click **Create invite**. WebGlazer emails the invite, and only that email address can accept it. If the email cannot be sent, WebGlazer shows the invite link so you can copy it and send it yourself. ### Share a link instead (Viewers only) For a Viewer, tick **Create a shareable link instead (Viewer only)** and click **Create link**. Anyone who opens the link joins as a Viewer of the selected projects. The link works once. Every invite, by email or by link, expires after 7 days. ## Accept an invite 1. Open the link from the invite email. 2. If you are not signed in, click **Continue with Google**, or **Use email instead** to sign in with email and password. 3. Check which company or brands you are joining and with which role, then click **Accept invite**. If the page says **This invite link is invalid or has expired**, ask a Company Admin to send a new one. ## See and manage your team **Settings** > **Members** lists everyone with access: each **Member** (email, status and whether MFA is on), their **Company role** and their **Projects**. Use **Search members** to filter, or the download button to export the list as CSV. Pending invites show an **Invited** badge. ![Settings, Members with search, Invite and three members with their roles](https://mintcdn.com/sia-webglazer/xTXVRghP1Bd03NXs/images/dashboard/mock-settings-members.webp?fit=max\&auto=format\&n=xTXVRghP1Bd03NXs\&q=85\&s=d987c40fcb265d759a445328562c0bdd) Illustration with sample data A **Company Admin** has these actions on each row: * **Manage access**: change the role and the projects in **Edit access**. * **Copy invite link**: for a pending invite. * **Suspend access** and **Restore access**: block someone for now without removing them. * **Revoke invite**: cancel an invite that was not accepted. * **Remove from company**: removes the person from every project of the company. They lose access at once. The owner's access cannot be changed or removed. Viewers and Members see only the people on their current project and cannot invite anyone. ## Leave a company If you were invited, click **Leave company** in **Settings** > **Members**. You lose access to every project of that company; its projects and billing do not change. To come back, a Company Admin has to invite you again. The owner cannot leave. To hand the company over, the owner clicks **Request ownership transfer** to pick a Company Admin, or uses **Contact support to transfer ownership**.