A Complete Guide to the AEOTrack Dashboard: Understand and Improve Your AI Visibility

Introduction

Search behavior is changing. Instead of typing a query into Google and scrolling through ten blue links, more and more people are simply asking an AI — ChatGPT, Claude, Gemini, Grok, or Perplexity — for a direct answer. When someone asks "what's a good project management tool for startups?" or "which analytics platforms do marketers recommend?", an AI model generates an answer on the spot, often naming specific brands.

This shift has created a new discipline, usually called answer engine optimization (the terms are explained here). Instead of optimizing purely for search engine rankings, it focuses on a different question: when an AI is asked a relevant question, does it mention your brand, and does it cite your website as a source?

AEOTrack is a platform built to help you answer that question with data instead of guesswork. It regularly asks AI models a set of real, relevant questions related to your industry and records whether your brand is named, how it's positioned relative to competitors, and whether your website is cited as a source behind the answer.

The AEOTrack Dashboard is where all of this comes together. It's the home base you'll return to again and again to check your AI visibility, see how it's trending, compare yourself with competitors, and figure out exactly what to fix next.

This guide is written for someone who has never opened AEOTrack before. By the end, you'll understand every major card, chart, and section on the dashboard — what it measures, how to read it, and what to actually do with the information.

---

What Is the AEOTrack Dashboard?

The AEOTrack Dashboard is the central control panel for monitoring your brand's presence inside AI-generated answers. Rather than tracking where your website ranks on a search results page, it tracks something more specific to the AI era: whether AI models name your brand when someone asks a relevant question, and whether they cite your website as the source behind that answer.

The dashboard is designed to help you answer questions like:

• How visible is my brand in AI-generated answers right now?
• How often is my brand mentioned compared with competitors?
• Which AI platforms mention my brand, and which ones don't?
• Is my visibility improving or declining over time?
• Which specific questions is my brand winning or losing?
• Where should I focus my next piece of content or website update?

Instead of a single vague "AI score," AEOTrack breaks visibility down into distinct, actionable pieces — so you always know not just how you're doing, but why, and what to do about it.

---

Dashboard Overview

The Dashboard reads from the top down: where you stand, anything to correct before you read a zero, the next thing to do, three tiles, and the detail behind them in a fold. What leads the page depends on how new your account and the website are. From the top:

1. The greeting and one line on where you stand, for example "Acme Coffee Roasters: AI names you on 6 of 10 questions." While the zero check below waits for your answer, the line only names the website.
2. The zero check, "Before you read this as a zero", when none of your latest answers names you but answers that cite your site write another name for your business. It comes above every zero on the page.
3. One line on what happened in your answers, when something did, for example "Bean Brothers Roastery was named in 26 answers to your questions, and you were not named in any."
4. The pages AI tools read for your questions, on the trial or in an account's first 21 days while your score is under 10: the pages read for the most of your questions in answers that did not name you, with the first one to go after marked.
5. The first screen, in an account's first 14 days, or while a website has no score history: your top question ("AI names A, B and C for ..."), the verdict across all your questions and Do this this week.
6. Is this you? and Check your setup: prompts that correct a setup the answers show is wrong. A website whose business was never confirmed also gets Check we have the right business here.
7. This week, once the first screen has gone: the first three open tasks of your plan.
8. Three tiles: Named in AI answers, with the Visibility score as its second number; Against tracked competitors; and Done for you this period. Under them are How this score works and How this score is calculated.
9. More detail: a fold with The other scores (the Branded figure, the Topics score, site readiness and share of voice), Score over time beside What changed, Where you rank beside Engine by engine, Recent citations, What you got this period and, from day 14, Where to spend your effort. Each link in the row under the tiles opens the fold at its section. Once the first screen has gone, Your result at a glance is the first thing in the fold.

Each of these is explained below, using a fictional example company: Acme Coffee Roasters, a small-batch roaster that sells beans and coffee subscriptions online. The screenshots come from demo workspaces of the same fictional business. The overview below shows its first day, with rivals such as Bean Brothers Roastery, Northwind Beans and Bluebird Coffee. The worked examples further down show it some weeks in (at acmecoffee.example), competing with Northwind Roasters, Bluebird Coffee, Harbor Beans and Ridgeline Coffee.

The top of the Dashboard on an account's first day: the zero check, what happened in the answers, the pages AI tools read for your questions, and the top question with who AI names for it.

The zero check

When none of your latest answers names you, but answers that cite or link your site write a name for your business that does not count yet, the Dashboard asks about that name before it shows any zero:

Before you read this as a zero: AI answers that cite your site also write the name "Acme Roastery Collective" (25 answers). Is this you?

Each name has Yes, add it and Not us. Yes, add it adds the name to the website's brand names, every stored answer is read again with it, and the page reads its numbers again: the zero may no longer be a zero. Not us is remembered for the website, so the name is not suggested again, here or in the Is this you? banner. A viewer, or an account that is read-only, sees the question without the buttons. Your first-result email opens with the same question. See Brand names and "Is this you?".

The first two weeks look different

For an account's first 14 days, and whenever a website has no score history yet, the Dashboard leads with one screen built from your answers rather than the full set of cards. It opens on your top question: who AI names for it, up to three names and then "and 2 more" (for example AI names Northwind Roasters, Bluebird Coffee and Harbor Beans for "best coffee subscription for home brewing"), how many of the engines named each, each engine's answer with the date it was asked, and Ask it yourself links (Ask ChatGPT, Ask Perplexity) that open the same question in that engine. Under it are the verdict across all your questions ("Not named in any of 45 answers yet" or "Named in 18 of 45 answers") with the line Visibility score N/100 · likely range ..., why you are not named for that question, Questions with your name in them, who was named across your questions, the sources they read, and Do this this week with See the full plan. When no question has an answer to show yet, the screen opens on the verdict instead. Your first 48 hours walks through it step by step.

From day 14, once the website has a score history, the first screen gives way to the layout described below. If you complete a plan task before then, the off-site plan card (for a website still at zero) and Where to spend your effort come back early; This week waits until the first screen has gone, so the page never lists your tasks twice. The first-result card (Your result at a glance) then sits under More detail. If you came from the free audit, it carries a bridge note that ties the audit's grade to the visibility score. For Acme Coffee Roasters it reads: "Your free audit graded your homepage B (72/100). That is site hygiene, not AI visibility. This score is whether AI names you: 18 of 45 answers. In your audit, ChatGPT named Northwind Roasters too."

Prompts that fix your setup

From the first check on, the Dashboard can ask you to correct something your answers show is set up wrong, so the numbers are about you. Is this you? is a banner of its own, on the Questions page too. The other three sit together in one section titled Check your setup, the first three of them shown and Show all for the rest; it appears once the first check has finished, to anyone who can edit the website. Each prompt shows its evidence, and nothing changes until you click:

• Is this you? A name the answers use for your business that does not count yet, such as one they link to your site. Add as a brand name counts it; Not us stops it being suggested. On a first result that names you nowhere, the names come first: "Before you read this as a zero". See Brand names and "Is this you?".
• Not a competitor? A tracked competitor that is a place, a directory, a marketplace or an everyday word, with Remove in one click. See Competitors that aren't competitors.
• Not you? Answers about a different business with your name, with Confirm who you are and new questions written for you. See Answers about a different business with your name.
• This question doesn't fit. A question whose answers can never name you, with a buyer question to replace it. The Questions page shows these too, under Check these questions. See A question that doesn't fit.

Check your setup with one Not a competitor? prompt and its evidence open: Yelp was tracked as a competitor, and the answers mention it only as a place to read reviews.

A prompt you keep (Keep tracking it, Keep this question, Dismiss) stays hidden unless its evidence changes, and the section says how many it hides ("1 check you chose to keep is hidden."). The Not a competitor?, Not you? and This question doesn't fit checks never email you; only your first-result email carries an Is this you? line.

---

Explain Every Dashboard Metric

Three tiles sit in one row under the first screen (once it has gone, under This week): Named in AI answers, with the Visibility score as its second number; Against tracked competitors; and Done for you this period. Each tile opens the page or the section behind it. The first tile has a ? beside its label that says what it counts and what the score is built from. The other four figures, Branded, Topics score, Site readiness and Share of voice, are under More detail, in The other scores. Here's what each one means.

Named in AI Answers

What it means: How many of your latest answers name you, out of the answers that could be read: the newest answer from each engine to each tracked question. It leads the first tile because it is the plainest fact on the page.

Example: On its first day Acme Coffee Roasters reads 0 of 35 answers: seven questions, five engines, and no answer that names it yet. Some weeks in, it tracks twelve questions and reads 32 of 60 answers. That count includes its two questions with its name in them and its one advice-only question, which the Visibility score leaves out.

How to interpret it: A zero on the first check is the usual starting point, not a fault; see Why AI does not name you yet. Before you read a zero, answer the zero check if the Dashboard shows one: the answers may already write your business under another name.

AI Visibility Score

What it means: The first tile's second number reads Visibility score. This is your overall composite score, out of 100, representing how visible your brand is across the AI questions AEOTrack tracks for you. Each engine's answer that names you scores 25 plus up to 75 for how prominently you are named, whether your own site is cited and the tone, an answer that only lists your own site as a source scores 22.5, and any other answer scores 0; the question scores are averaged across engines and recency-weighted into this number. Questions no engine answers with any business are advice-only and left out. Since scoring version 9, questions that contain your brand are left out too and scored on their own (see "Branded Score" below). A failed answer never counts as "not named": the engine's last good answer stands in for it.

Under the two numbers, one line gives the sample the score stands on: its ± band, how many answers, how many engines and the date of the check, for example "±4 · 40 answers · 5 engines · checked 12 Sep 2026". The count above it covers every tracked question, while the score leaves out advice-only questions and questions with your name in them. When that makes the two counts differ, the line says which answers it means, for example "±19 · 45 answers in the score · 5 engines · checked 29 Sep 2026". In your first 14 days the first screen prints the same facts under its verdict (Visibility score 0/100 · likely range 0 to 3 · 35 answers · 5 engines · checked 23 Oct 2026), so the tile leaves the line out. The ? beside the tile's label gives the rates behind the number, how many advice-only questions were left out, and that questions with your name in them are scored apart. How this score works, under the tiles, opens the short version of the formula; the full version is in How the score works. A score with no check in 30 days is marked stale.

The How this score works panel, opened from the link under the tiles.

Example: Some weeks in, Acme Coffee Roasters' Visibility score is 38/100, up 5.2 points since the last check. The tile counts 60 answers, and the line under it reads "±19 · 45 answers in the score · 5 engines · checked 29 Sep 2026": the score leaves out the 10 answers to its two questions with its name in them and the 5 answers to its one advice-only question, and the ? says so.

How to interpret it: A low score (under 20–30) generally means AI models rarely name your brand when relevant questions are asked. A rising score suggests your visibility is improving; a falling score suggests you're losing ground — either because competitors are being mentioned more, or because AI models simply aren't finding enough about you to reference.

Why it matters: This is your Questions score, the headline number. It's the fastest way to know, at a glance, whether your AI visibility strategy is working.

What you should do: Don't obsess over small day-to-day movements in isolation — check the trend over several checks first (see "Score Over Time" below) before drawing conclusions.

Against Tracked Competitors

What it means: Your place among the brands you track, by share of voice, with the gap to the nearest one in mentions, for example "4th of 5 tracked" and "Bluebird Coffee is next above you: 194 mentions to your 183". It is a rank among the competitors in your account, not among every brand AI names, which is why it says "tracked". The tile opens the Share of voice page.

How to interpret it: Until any tracked brand has been named, the tile says "No tracked brand has been named yet" rather than ranking a race that has not started. A place, a directory or an everyday word in your competitor list distorts the rank; the Not a competitor? prompt removes one.

Done for You This Period

What it means: How many answers AEOTrack read for you in the period, and about how long that would take by hand at five minutes an answer, for example "35 answers read" and "≈ 2.9 hours of manual checking". The tile opens What you got this period under More detail.

Branded Score

What it means: Your score on the questions that contain your brand name: branded questions such as "is Acme Coffee Roasters any good", and near-brand questions that contain your name once the spaces are removed. Any engine names you when it is asked about you by name, so this figure shows whether AI knows you and how it describes you, not whether it recommends you to buyers who have not heard of you yet.

Since scoring version 9 (October 2026) these questions are left out of the AI visibility headline and scored apart, so the headline is your score on the questions buyers ask before they know your name. Their own score is the figure labelled Branded: on the Dashboard under More detail, in The other scores, and on the Questions page beside the visibility score. In your first 14 days the first screen also shows it under Questions with your name in them. It is not part of the visibility score, and it shows no number while none of your questions contains your name. If every question you track contains your name, the headline counts them all, so you still have a number, and the line under it says so: "Every question you track contains your name, so this score counts them." See Branded and near-brand questions.

Example: Acme Coffee Roasters tracks two questions with its name in them, "is Acme Coffee Roasters any good" and "Acme Coffee Roasters reviews". Its Branded figure is 60/100, while its AI visibility headline, built from its other questions, is 38/100.

How to interpret it: A high Branded figure beside a low headline means the engines know you when asked by name but do not yet put you forward when a buyer asks the category question. A low Branded figure means that even when asked about you by name, the engines say little about you or point to other businesses.

Why it matters: It keeps reputation apart from discovery. Counted in the headline, branded questions would flatter it: any engine names you when your name is in the question.

What you should do: Keep a branded question or two to watch how AI describes you, and put most of your questions on what buyers ask before they know your name, which is what the headline measures.

Topics Score

What it means: This tracks your visibility in the broader topic overviews the same AI engines write when asked about a subject rather than a specific buying question. It is measured over your tracked topics across ChatGPT, Gemini, Claude, Perplexity and Grok; Google's AI Overviews are not part of it. (Older screens call it the GEO score.)

Example: Acme Coffee Roasters has a Topics score of 34/100.

How to interpret it: This score can move independently of your standalone AI-assistant visibility, since generative search summaries pull from different signals than a chat-based AI model.

Why it matters: Because generative search summaries appear directly inside familiar search engines, they can be seen by a very large volume of everyday searchers — even people who never use a dedicated AI chatbot.

What you should do: If your Topics score lags behind your Questions score, look at how well your existing SEO-optimized content already answers common questions clearly and directly, since generative search summaries often lean on well-structured, authoritative web content.

Site Readiness Score

What it means: This score evaluates how "AI-crawlable" and citation-friendly your actual website is — things like whether your content is structured clearly, whether pages have enough substantive detail, and whether AI systems can easily extract accurate information from your site.

Example: Acme Coffee Roasters scores 78/100 on site readiness (the Website Health figure on the Pages audit, across 18 scanned pages).

How to interpret it: A high site readiness score (like 90/100 or more) means the technical and structural groundwork is solid — your site itself isn't the bottleneck. A middling score like Acme's 78 means crawlers can reach the site but some pages still hold it back: in its case, thin pages and missing FAQ schema. A low score would suggest technical or content-depth issues are actively preventing AI systems from reading and citing your pages.

Why it matters: A well-optimized, easy-to-crawl site is the foundation that makes everything else possible. Even great content can go unnoticed if AI systems struggle to parse or trust the page it lives on.

What you should do: If site readiness is already high, know that your visibility gap is unlikely to be a technical problem — it's more likely about topical coverage, content depth, or the fact that AI simply doesn't have enough reason yet to cite you over competitors.

Share of Voice

What it means: Share of voice shows what percentage of all brand mentions (across your tracked competitors, combined) belong to you, and your numeric rank among those competitors.

Example: Acme Coffee Roasters holds a 20% share of voice, ranking #4 of 5 tracked brands. The top competitor, Harbor Beans, holds 22%.

How to interpret it: If a leading competitor is mentioned four times more often than you, that tells you exactly how large the visibility gap is in relative terms — not just in the abstract. Acme's gap is small: #4 of 5 sounds low, but it is 2 points behind the leader in a close field.

Why it matters: Share of voice puts your visibility score into competitive context. A visibility score alone doesn't tell you if you're behind an entire industry or ahead of it — share of voice does.

What you should do: Use the Where you rank leaderboard under More detail (see "Where You Rank," below) to see exactly which competitors are outperforming you and by how much.

---

"Where You Rank": Competitor Leaderboard

Under More detail, AEOTrack displays a "Where you rank" panel — a straightforward leaderboard of every brand it's tracking for your account, ranked by number of mentions and share of voice.

What it is

A ranked table listing each tracked brand (including yours), the total number of times each was mentioned across all monitored AI answers, and the resulting share-of-voice percentage.

What it measures

Raw mention volume and relative share of voice, side by side, for every competitor in your tracked set.

Why it matters

This table answers a simple but important question: "Compared to who, exactly, am I behind — or ahead of?" It transforms an abstract score into a concrete competitive picture.

How to interpret it

Example table (fictional data):

In this fictional example, Acme Coffee Roasters sits in fourth place of five. The top four are within 26 mentions of each other, so a few more answers naming Acme would move it up. The Benchmarks page shows the same split: 183 of 520 collected answers named Acme.

What action to take

If a single competitor dominates share of voice, investigate what content or positioning is likely earning them that volume of mentions — then look for the specific questions where they're winning and you're not (see the "Question Analysis" and "This Week" sections below).

---

AI Platform / Model Performance ("Engine by Engine")

Next to the leaderboard, AEOTrack shows an "Engine by engine" panel — a breakdown of your visibility performance separated out by individual AI platform.

What it is

A row-by-row list of AI platforms (for example, ChatGPT, Claude, Gemini, Grok, and Perplexity), showing how often each platform named your brand and what rank you held on that specific engine. It counts every answer collected in the last 90 days, not just the latest check, and the card says how many; the score at the top of the page reads the latest check only, so the two can differ. Each percentage carries its count (for example 3/12), because an engine with only a handful of answers moves by a lot per answer.

What it measures

Platform-specific visibility. Rather than a single blended score, this shows exactly which AI systems know about and cite your brand — and which ones essentially don't.

Why it matters

Visibility can vary dramatically from one AI platform to another. Each AI system has its own training data, its own approach to web browsing and citation, and its own audience. A brand that performs strongly on one platform may be almost invisible on another.

Example (fictional data):

In this fictional example, Acme Coffee Roasters is named on every engine, but not evenly. It leads on Perplexity and is weakest on Claude, where the Share of voice page also shows it well behind Harbor Beans and Northwind Roasters. The rank depends on how often rivals are named on the same engine, so a higher named rate (ChatGPT) can still rank lower than a smaller one (Grok). On Claude, Acme's pages were read 7 times on questions where the answer still did not name the brand ("read 7×, not named" means the site's content was accessed as a source but the AI didn't name the brand in its final answer). The five counts add up to the 183 of 520 answers that named Acme across all checks.

How to interpret it

A "read, not named" label is a useful diagnostic: it tells you AI systems are aware your content exists and are pulling it in as a potential source, but aren't yet confident enough — or don't see clear enough signals — to actually recommend your brand by name. This is different from a platform that never reads your site at all.

What action to take

Prioritize investigating platforms where you're being "read but not named" — the engine already reads your content there, so the gap is in being named, not in being found. Compare what competitors are doing on those specific platforms to understand the gap.

Avoid drawing conclusions about why a specific AI platform behaves the way it does — the internal ranking or selection logic of any given AI model isn't publicly documented, so it's best to treat these differences as observed patterns to investigate rather than confirmed technical explanations.

---

Score Over Time

Under More detail, the Score over time chart tracks how your visibility score has moved across your tracked check-ins, with the What changed panel beside it.

What the chart shows

• X-axis: Dates of each AEOTrack check, spanning your selected window.
• Y-axis: Your AI Visibility Score (0–100) at each check.
• Line: Your score plotted across each check-in date.
• Shaded band: The range to expect between two checks, drawn around each measurement. Because AI models can give slightly different answers to the same question from one run to the next, a single check carries some natural variability. A movement that stays within the shaded band should be read as statistical noise rather than a confirmed trend.
• Version marks: When the scoring version changes, the stored answers are re-read once and the chart marks the step with a note saying what changed. Version 9 (October 2026), for example, moved branded questions out of the headline. A step at a mark is the re-reading, not a change in how often you are named.

Example (fictional data): Acme Coffee Roasters ran 16 checks from 04 Aug 2026 to 29 Sep 2026. Its score climbed from under 10 to 38, and the chart labels the change +30.1. The line rises steadily and ends well above where the shaded band started, so this is real progress, not noise. Had the next check come in at 34 instead, that dip would sit inside the ±19 band and should be read as noise, not as a decline.

What each element means for you

• A line trending clearly upward, outside the shaded band, indicates a real, meaningful improvement in visibility.
• A line trending downward, outside the shaded band, indicates a real decline worth investigating.
• Movement that stays inside the shaded band is most likely random variation between AI answer runs, not a genuine change.

"What Changed" Panel

Next to the trend chart, AEOTrack includes a "What changed" panel that summarizes specific shifts across a stated window. The panel names the window in its heading (for example "since 21 Sep" or "last 7 days"), so you always know which two checks are being compared, and it only lists movements that cleared the ± range. For example:

• A certain number of questions that newly became a "stronghold" (where you're now consistently named and cited).
• Questions that are now being read by AI but still not resulting in a mention.
• Questions where you used to be mentioned but are no longer cited.
• Questions where you remain completely invisible.

Example (fictional data): Acme Coffee Roasters' panel covers "Between 19 Aug 2026 and 28 Sep 2026, last 30 days" and lists "+6 questions now a stronghold", "-1 question mentioned, but not cited" and "-5 questions still invisible".

How to use this section

• Use a short window (a handful of days) to catch sudden, recent shifts — useful right after a content or site change.
• Use a longer historical window to evaluate whether a strategy is working over weeks or months, filtering out day-to-day noise.
• Always check whether a change falls inside or outside the shaded range (the range to expect between two checks) before treating it as a real trend.

---

Citations vs. Mentions: "Where to Spend Your Effort"

One of the most practically useful sections on the AEOTrack dashboard is the "Where to spend your effort" panel under More detail (from day 14, or sooner once you complete a plan task), which cross-tabulates two related but distinct concepts: being mentioned by an AI, and being cited as a source.

What a mention is

A mention means an AI model named your brand somewhere in its answer to a question.

What a citation is

A citation means the AI model referenced your actual website as a source it drew information from — sometimes with a visible link, sometimes just as an underlying reference.

Why the difference matters

Being mentioned and being cited are not the same thing, and the gap between them tells you very different things:

• "Stronghold" — AI both names you and cites your website. This is your strongest position; the goal is to defend it.
• "Visible, not cited" — AI recommends your brand, likely from its own training knowledge, but doesn't link to your site as a source. This can be a fragile kind of visibility, since it isn't reinforced by your own web content.
• "Cited, not visible" — AI reads your page as a source but stops short of naming your brand in its final answer. This usually signals a content or positioning issue rather than a discovery issue.
• "Needs attention" — Neither named nor cited. You're effectively invisible on this specific question.

Example (fictional data): Acme Coffee Roasters has 5 "stronghold" questions, 2 "visible, not cited", 1 "cited, not visible" and 5 "needs attention", out of 13 questions. The panel counts across 1,040 AI responses from the last 12 months, a longer span than the score at the top of the page.

What action to take

• For "cited, not visible" questions, review the page AI is reading — it may lack a clear, direct statement of what your product does and why it's relevant to the question.
• For "visible, not cited" questions, make sure you have a strong, citable page that directly and clearly answers that specific question, so future citations reinforce the mention.
• For "needs attention" questions, start with the sites the engines read for that question: getting mentioned there is usually the quickest way in, since there's currently no visibility at all to build on. See Get named: work the off-site plan.

---

Question Analysis

AEOTrack tracks visibility on a set of specific questions: real, relevant things a prospective customer might plausibly ask an AI system.

What a tracked question is

One of the questions used to test your visibility, word for word, for example "best coffee subscription for home brewing" or "which coffee roaster has the freshest beans" (fictional).

Why questions matter

Questions represent the real language customers actually use when they ask AI for recommendations. Tracking visibility question by question, rather than just an aggregate score, shows you exactly which questions you're winning, and which ones you're losing, and to whom.

How to identify weak-performing questions

In your first 14 days the first screen leads with one of them, your top question, and says who was named instead and why you were not. After that, Your result at a glance, at the top of More detail, lists the latest check per question, and the Questions page lists every question with who was named.

Example (fictional data): For the question "best light roast coffee brands," AI names Bluebird Coffee and Northwind Roasters, but never Acme Coffee Roasters. On "best gifts for coffee lovers" the Dashboard row reads "2 rivals named, not you (5 answers)".

How to prioritize

Not all questions carry equal weight. Start with the questions a competitor is winning, and open Sources for them: the pages the engines read when they named the competitor are your targets. Get named: work the off-site plan walks through it.

---

"This Week": The Next Three Tasks

Once the first screen has gone (from day 14, once the website has a score history), a This week card sits above the tiles: the first three open tasks of your action plan, in the plan's order. Until then its place is taken by Do this this week at the foot of the first screen: one task, with See the full plan. Completing a task does not bring This week back early, so the page never lists your tasks twice.

What it is

The first three open tasks of the plan, in the plan's order. The plan has three sections: Get onto the sites AI reads first, then Move up in the answers you're already in, then a collapsed Hygiene checklist. So the top tasks are usually sites the engines read while answering your questions without naming you. Tick a task to mark it done in the plan itself, and the next open task takes its place. Open on a row goes to that task in the plan, and the card's link opens the whole plan. The "to do this week" count in the top bar counts the same tasks.

Example items (fictional, illustrative)

1. "Get onto roasterratings.example" — read 71 times in 30 days for "best light roast coffee brands" while the engines named Bluebird Coffee and Northwind Roasters, never Acme Coffee Roasters.
2. "Get onto giftguide.example" — read 32 times for "best gifts for coffee lovers", where Ridgeline Coffee and Bluebird Coffee were named.
3. "Add statistics, quotes and sources to the pages AI already reads" — Acme's own pages the engines read, and for which questions.

How to interpret it

The sites are ordered by how often the engines read them, weighted by engine, with reads in answers that named a competitor counting in full. Answer tasks follow, and housekeeping comes last. Each row shows its evidence where the task has a count, for example "Read 71 times in 30 days", and no label otherwise. How to work each kind of site is in Get named: work the off-site plan.

What action to take

Work through the list from top to bottom. After a change (getting listed on a site, adding detail to a page the engines read), watch Score over time and the What changed panel under More detail on your next check to see whether the action produced a measurable, outside-the-noise-band shift.

---

How to Read the Dashboard: Step-by-Step

Here's a practical workflow for reviewing your AEOTrack dashboard each time you check in:

1. Answer anything above the numbers first — The zero check ("Before you read this as a zero"), Is this you? and Check your setup. A name of yours that does not count, or a place in your competitor list, makes every figure below it wrong.
2. Read where you stand — In your first 14 days, who AI names for your top question and why you were not named; after that, Named in AI answers and the Visibility score on the first tile, and whether the change cleared the ± range.
3. Review the trend chart — Open Score over time under More detail and determine whether any movement is a real trend or stays inside the range to expect between two checks.
4. Check your share of voice and ranking — Against tracked competitors gives your place; the Where you rank table under More detail gives the whole field in concrete terms.
5. Review engine-by-engine performance — Identify which AI platforms know about you and which ones don't, and note any "read, not named" opportunities.
6. Analyze mentions vs. citations — Use the "Where to spend your effort" panel to see how many questions are strongholds versus needing attention.
7. Investigate specific questions — Look at which individual questions you're losing, and to which competitors.
8. Do this week's tasks — Start This week (in your first 14 days, Do this this week) with the top site AI reads.
9. Take action and re-check — Make the recommended change, then return on your next scheduled check to measure the effect.

---

Example: Understanding a Fictional Dashboard

Let's walk through a complete fictional scenario using Acme Coffee Roasters.

Maya logs into Acme Coffee Roasters' AEOTrack dashboard and sees:

• Named in AI answers: 32 of 60 answers in the latest check, with a Visibility score of 38/100 (±19, from the 45 answers in the score), up 5.2 points since the previous check
• Against tracked competitors: 4th of 5 tracked; a 20% share of voice, with Harbor Beans on top at 22%
• Branded: 60/100 on its two questions with its name in them, apart from the headline
• Engine breakdown: Named on all five engines, strongest on Perplexity and weakest on Claude
• Trend chart: 16 checks from 4 Aug to 29 Sep 2026, up 30.1 points, well outside the range
• Mentions vs. citations: 5 stronghold questions, 2 visible-but-not-cited, 1 cited-but-not-visible, and 5 needing attention
• This week: first, "Get onto roasterratings.example": read 71 times in 30 days for "best light roast coffee brands", where the engines named Bluebird Coffee and Northwind Roasters

Interpretation: Acme Coffee Roasters is making real progress and is close to the leaders: its share of voice is only 2 points behind Harbor Beans. The site is in reasonable shape (78/100 site readiness, per the earlier example), and crawlers can reach it, so the remaining gap is off its site. Its weakest questions, "best light roast coffee brands", "best decaf coffee beans" and "best gifts for coffee lovers", are answered from review sites and list articles that name its rivals and not Acme. The clearest next step is the top action: get Acme onto roasterratings.example, then the next two sites, and watch those questions on the next checks.

(All figures above are illustrative examples only and do not represent real AEOTrack customer data.)

---

How to Use Dashboard Insights to Improve AI Visibility

Once you've identified gaps using the dashboard, here are general categories of improvement to consider:

• Get mentioned on the sites the engines read. Review sites, "best X" list articles, Reddit threads, YouTube videos and press are where most AI answers get their brand names. This is the strongest lever; see Get named: work the off-site plan.
• Create genuinely detailed, useful content. Thin pages with only a sentence or two typically give AI little to work with.
• Directly answer the specific questions you track. If a tracked question asks "what's the best tool for X," make sure a page on your site clearly and directly answers that exact question.
• Build clear, unambiguous entity information. Make it easy for AI (and readers) to understand exactly what your product does and who it's for.
• Address competitor content gaps. Where a competitor is named on a question and you aren't, look at what they've published, and where they are mentioned, that you aren't.
• Strengthen the pages AI already reads. Add statistics, quotes and sources to them; the plan's second section lists which ones.
• Monitor consistently. AI visibility shifts gradually — regular check-ins matter more than any single measurement.

No individual tactic guarantees that an AI platform will mention your brand; these are directional best practices based on the kinds of gaps the dashboard identifies, not guarantees of outcome.

---

Common Dashboard Scenarios

Scenario 1: Visibility is increasing
Identify which specific questions or platforms are driving the improvement, and continue reinforcing that content so the gain holds.

Scenario 2: Visibility suddenly drops
Check the "What changed" panel first to see which specific questions moved, and confirm the drop is outside the range before making major changes.

Scenario 3: Mentions increase but citations decrease
This can mean AI is starting to recognize your brand generally, but is relying on general knowledge rather than your own website — strengthen the specific pages relevant to those questions.

Scenario 4: Competitors consistently outperform you
Use the "Where you rank" and question-level views to identify exactly which questions they're winning, then look at the sources the engines read for those questions and get onto them.

Scenario 5: Strong website traffic but weak AI visibility
Traditional search performance and AI visibility are measured differently and don't always move together — a page can rank well in conventional search while still not being cited or named by an AI model.

Scenario 6: One AI platform performs much better than another
Focus on platforms where you're being read but not named, since discovery is already happening there — and avoid assuming specific reasons for the gap without more investigation, since each platform's internal selection process isn't publicly documented.

---

Dashboard Best Practices

• Monitor trends over multiple checks rather than reacting to a single data point.
• Compare consistent time windows when evaluating progress.
• Pay attention to the range before calling something a "real" change.
• Track specific questions you care about, not just the aggregate score.
• Review competitor movement regularly, not just your own numbers.
• Look at mentions and citations together, not in isolation.
• Work the action plan from the top rather than guessing.
• Treat the dashboard as an ongoing monitoring tool, not a one-time report.

---

Frequently Asked Questions

What is the AEOTrack Dashboard?
It's the central monitoring hub where you can see how visible your brand is in AI-generated answers, how that compares to competitors, and what to improve next.

What does AI visibility mean?
It refers to how often and how prominently AI models mention and cite your brand when answering questions relevant to your industry.

What is an AI brand mention?
A mention is any instance where an AI model names your brand in its response to a tracked question.

What's the difference between a mention and a citation?
A mention means your brand was named. A citation means the AI referenced your actual website as a source behind that answer. You can have one without the other.

What does "share of voice" mean?
It's the percentage of total mentions (across all tracked competitors) that belong to your brand, along with your rank among them.

Why is my visibility different across AI platforms?
Each AI system has different training data and different approaches to sourcing and citing information, so visibility can vary meaningfully from one platform to another.

Why did my visibility score change?
Check the "What changed" panel and the shaded range on the trend chart — some movement is normal variability rather than a confirmed shift.

How often should I check the dashboard?
Regularly enough to catch real trends, but frequently checking small day-to-day movements is less useful than reviewing broader trends across several checks.

How should I compare my brand with competitors?
Use the "Where you rank" leaderboard for overall standing, and the engine-by-engine and question-level views for more specific comparisons.

What should I do if competitors have stronger AI visibility?
Identify the specific questions they're winning and the sites the engines read for them, then get your brand onto those sites. Get named: work the off-site plan explains how.

What does "site readiness" measure?
It reflects how well-structured and easy to reference your website is for AI systems — separate from whether AI actually chooses to mention or cite you.

What is a "stronghold" question?
A tracked question where AI both names your brand and cites your website — your strongest, most defensible position.

Why does the Dashboard say "Before you read this as a zero"?
None of your latest answers names you, but answers that cite your site write another name for your business. If the name is yours, Yes, add it counts it and the numbers are read again; if not, Not us stops it being asked. See "The zero check" above.

Why does the Dashboard ask "Is this you?"
The answers write a name for your business that does not count as naming you yet, for example one they link to your site. Adding it re-reads every stored answer with it. See Brand names and "Is this you?".

A town or a directory is in "Where you rank". Why?
It is tracked as a competitor, so every answer that mentions it counts for it. The Not a competitor? prompt removes it in one click; see Competitors that aren't competitors.

---

Conclusion

The AEOTrack Dashboard isn't meant to be a one-time scorecard — it's an ongoing monitoring and decision-making tool. AI visibility shifts gradually, shaped by content, site structure, and how each AI platform discovers and evaluates information over time.

To get the most out of it:

1. Monitor your overall visibility regularly.
2. Understand trends rather than reacting to single data points.
3. Review mentions and citations together, not in isolation.
4. Compare yourself against tracked competitors.
5. Investigate specific questions where you're losing ground.
6. Identify concrete content opportunities.
7. Track improvement over time, one check at a time.

If you're ready to see exactly where your brand stands in AI-generated answers, log in to your AEOTrack dashboard and start at the top: answer the zero check or Check your setup if either is there, then take the first task under This week (in your first 14 days, Do this this week), which usually starts with the sites AI reads when it doesn't name you.