Brand Dashboard

The Brand Dashboard is the property home: a grid of cards, one per metric, each showing a single number and how it moved over the last seven days. Every card opens a drill-down page that explains the number, and every drill-down page leads to the evidence underneath it — the actual AI answer text, the individual check, the individual citation.

That three-level structure is the whole idea. You should never have to take a metric on faith. If Share of Voice fell four points this week, you can get from the card to the prompts that produced it and read what the assistant actually said, without leaving the property.

What You'll Learn

• The three levels — headline KPI, drill-down, raw data — and how to move between them
• What each module measures and what its drill-down gives you
• How the dashboard changes for a person subject
• How annotations put your actions on the charts
• How to switch the dashboard on, and how to get back to the classic views

The Three Levels

L1 — the dashboard cards

Each card carries one headline number, a delta badge comparing the last 7 days to the 7 before them, and a short caption. Cards do not compete for attention: the number is the point, and everything else on the card exists to make it interpretable.

Every card title has an info icon. Hovering it opens the metric's definition — what is counted, what is excluded, which direction is good — with a link through to the documentation for that metric. Read those before you act on a number for the first time.

A card whose metric isn't set up yet shows a setup prompt instead of an empty chart. "Add competitors to see how your Share of Voice compares" with a link to the competitors page is more useful than a zero.

Above the grid, a freshness line states what the numbers rest on: how many prompts ran, across how many platforms, and when they were last measured. If the property has no active project, a band appears explaining that daily tracking is off and these numbers only update when you run a report — with a link to set a project up. If nothing has ever been measured, the whole dashboard is replaced by a single prompt to run your first AI Visibility Report.

L2 — the routed drill-down pages

Clicking through from a card lands on a real page with its own URL, so you can bookmark it, share it with a client, or export from it.

Drill-downs open on the same 7-day window the card showed, with 30- and 90-day presets in the time-range selector. Changing the window rewrites the URL, so browser-back after a detour into the raw data returns you to the range you had chosen rather than snapping back to 7 days.

Two drill-downs deviate deliberately. SEO opportunities defaults to 90 days, because grounding-gap runs happen weekly and a 7-day window would show a single run and no movement at all. Site readiness has no window selector, because audits are sparse events you trigger rather than a daily series.

Each drill-down carries the standard export control: the underlying rows as CSV, or the chart as an image.

L3 — the raw data

The bottom level is the evidence, and it is always reachable.

• Answer records. From the visibility, consistency and locations drill-downs, a View answers button opens the answers for any measured day — the full response text an assistant produced for a tracked prompt, on its own routed page.
• Individual checks. A single stored brand-consistency or identity-accuracy check, in full: every scored field, per platform, with and without web browsing, alongside the answer that was scored.
• Individual runs. Every stored run for one grounding search, with the rank axis reversed so #1 sits at the top and gaps in the line mean you really didn't rank that week.
• Individual citations. The earned-media drill-down hands off to Citation Intelligence, where every citation can be expanded to its source page, citation type, author and publish date.

Turning It On

The Brand Dashboard is opt-in, per user. Your choice follows you across properties and devices; it does not change what your teammates see.

There are three ways in:

1. From the banner. An invitation banner sits at the top of the classic property dashboard for eligible properties. Preview opens the dashboard without committing to anything, Make it my default switches your property home over, and the dismiss button hides the banner for good.
2. From account settings. Settings → Account has a checkbox for the Brand Dashboard. Toggle it there at any time.
3. From the URL. Append ?dashboard=new to a property URL to preview, or ?dashboard=classic to force the classic dashboard for that visit.

While previewing, a ribbon above the grid offers Make it my default or Exit preview. Once you've opted in, that ribbon collapses to a quiet Switch back to classic dashboard link in the top-right.

The classic views are still there. Switching your property home does not remove or rewrite anything: Citation Intelligence, the AI Readiness Audit, Projects, reports and every other page keep their own URLs, their own filters and their own time windows. The drill-down pages are additions, not replacements, and the numbers you already report on don't move.

Brand, product and person properties are eligible. Category and location sub-properties always render the classic dashboard.

The Modules

Cards render in a fixed order, and a card that doesn't apply to the property is not rendered at all rather than shown empty.

Share of Voice

Of all the AI answers to your tracked consideration- and decision-stage prompts, the share that mentioned your brand. Brand-identity and comparison prompts are excluded — they have their own modules. The card pairs the percentage with the raw brand-mention count and a 14-day trend line on a fixed 0–100% scale.

The drill-down keeps that headline pinned at the top and then takes it apart: per-platform chips for ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews (click one to scope the whole page to it), the buyer's-journey band showing awareness, consideration and decision with a bracket making "Share of Voice = consideration + decision" explicit, and a trend chart with an optional 90% confidence range. The awareness stage also carries Share of Influence — how much of your key-message story is pulling through. Filters for time range, project, platform and prompt tag apply to everything on the page.

When the brand has locations, this card is measured brand-wide and says so; per-market numbers live in the Locations module.

Competitors

Each tracked competitor's share of voice across the same prompts, so the numbers are directly comparable, with your own row in the list and your rank out of the full set. Rows are toned by the gap to you. Competitors with brand-comparison prompts running get a Compare link.

The drill-down gives the full ranked list plus competitor share of voice over time, with a selector that narrows both the rows and the chart to you-versus-them. Compare opens a head-to-head page: favored-rate trend, what answers favor each side for, strengths and weaknesses, and the raw answers behind them. See Competitors for how the set is managed.

Locations

Rendered only when the brand has location children. Each market's latest share of voice as a ranked bar list.

The drill-down is the cross-market comparison: one trend line per market with the location-neutral brand-wide run as a dashed reference, plus a table of each market's latest measurement. Opening a market gives you its own funnel, its own prompts and the full answers for any day. See Locations.

Brand Consistency

How similarly AI platforms describe your brand across runs, scored 0–100. When brand-identity prompts are running, the card also shows message drift — how far today's story has moved from your established themes. Lower drift is better, and movement inside ±0.05 is treated as noise rather than a trend.

The drill-down carries per-platform score trends, expected-message coverage, head-to-head favored share, answer themes and attribute mentions, and a list of recent checks. Each check opens in full at L3. When every check in the window came from a report rather than daily tracking, the chart switches from a line to points and says so, instead of implying continuity between two dots six weeks apart.

For a person subject this slot is replaced by Identity Accuracy — see The Person Variant below.

Key Messages

How often each message you want AI to repeat — and each claim you'd rather it didn't — actually surfaces in tracked answers. Matching is semantic, so paraphrases count. The card shows the number of tracked messages, the biggest gain and biggest drop, and the top movers as sparklines. Negative-sentiment messages invert the delta badge: a rising negative claim reads as bad news.

The drill-down plots per-message pull-through over time, with tag and sentiment filters and the ability to pin up to four messages onto one chart for comparison. See Key Messages.

Citations and Goal Impact

One card with two faces, chosen by whether you have goals to measure against.

• Goal impact appears when any active project has goals: the share of tracked AI citations driven by the pages and placements those goals target. This is the direct proof that content and PR work is moving AI answers. Its drill-down shows the property-wide rollup and then every active project with goals — influence share, attributed mentions, near-misses — each linking into the project's own pages.
• Citations is the fallback: the owned-versus-earned split of your AI citations, shown as a ratio and a bar. It is deliberately toneless, because neither direction is unambiguously good. It drills into Citation Intelligence.

Set up goals in Projects to switch this card from describing your citation mix to measuring your work.

SEO Opportunities

The grounding searches AI ran to ground its answers where competitors ranked and you didn't, counted from the most recent grounding-gap run, with the top three by usage listed. These runs are weekly.

The drill-down is the full opportunity workspace: a KPI row (searches tracked, open gaps, high-impact gaps, movers, average impact), rank-band distribution over the window, a moving queries list showing which searches changed band since their own previous run, and a filterable, paginated table of every tracked search. Each row where ChatGPT produced the fan-out carries an Optimize button that opens the Citation Optimizer with that exact query already selected as a verified fan-out. Clicking a row opens its full run history. Related reading: AI Visibility Rankings.

Site Readiness

The overall score from your most recent completed site audit, with the change against the audit before it, and your three weakest dimensions listed underneath.

The drill-down turns the latest audit into a cross-page work list: every finding, filterable and paginated, what each dimension has done over the last twelve audits, and a re-check action for the pages you've fixed — closing the find → fix → re-audit → verify loop. The AI Readiness Audit manager keeps the launcher, per-page detail and chunk analysis; this page links out to it rather than duplicating it.

Earned Media

Third-party sources AI already trusts that mention your competitors but not you — a PR pitch list assembled straight from citation data, with the top outlets and their citation counts.

The drill-down ranks those outlets by placement value (AIPVS), groups them into gap categories, opens a per-publisher drawer, and lets you select a set of domains and hand them off in bulk into a PR project. Its raw-data level is Citation Intelligence. See also Publisher Lookup and Coverage Groups.

AI Traffic

Real visits measured on your own site over the last 7 days: AI agents reading your pages, and humans clicking through from AI answers. Each with a sparkline and a period-over-period trend.

This is the only module that depends on the tracking script rather than on prompt runs. An unverified property shows a setup prompt instead of numbers. See AI Traffic Analytics.

Annotations on the Charts

Every time-series chart in the drill-downs carries diamond markers along the bottom, one per annotation in the visible window. Hovering a marker shows the date and the title you gave it.

This is what turns a step in a trend line into something with a cause attached. "Published the CRM comparison guide" sitting three days before a citation climb is the difference between a chart and an argument. Annotations are property-scoped, so every event on the property appears on every drill-down chart regardless of which project it belongs to; an annotation outside the selected range simply isn't drawn, so widen the window to bring it back.

Project goal achievements create annotations automatically. Everything else — content publishes, technical changes, competitor launches, platform updates — you add yourself.

The Person Variant

A person property gets the same dashboard composed from a different set of modules. The changes are structural, not cosmetic: modules that depend on owning a crawlable domain are absent, because a person doesn't have one.

Identity Accuracy replaces Brand Consistency. Instead of "how similarly does AI describe this brand", the metric is "how accurately does AI describe this person" — role, employer, biography — scored against neutral primary sources. The card breaks the score out per model and flags the two failure modes that matter to an individual: confusion with a same-name person, and facts that have gone stale. It drills into the same consistency pages a brand's score does, because identity checks are stored in the same rows.

The swap is a swap, not a relabel. The Brand Consistency card's message-drift half depends on brand-identity prompts, which a person never has, so showing it would offer a setup step they cannot take.

Peers replace competitors. Same math, same drill-down, different name for the set — and peers are added by name, with no website needed. Fellow executives, industry voices, people you're frequently co-mentioned with.

Earned media stays, reading as the outlets citing your peers and not you — podcasts, expert roundups, trade press. The mention scan is a name-and-alias text match, so it never needed a domain in the first place.

Four modules are hidden:

• SEO opportunities and Site readiness, which both assume a crawlable site you own.
• AI Traffic, which requires the tracker installed on your own domain.
• Locations, which is a geographic slice of a brand.

The Citations card only appears with goals. Its fallback reading is the owned-versus-earned split, and a person has no owned domain, so that ratio would be 0:100 forever — a fact about the subject type, not about visibility. A person's dashboard therefore skips that reading entirely: the card appears only when an active project has goals, showing goal impact, and stays hidden otherwise.

The opt-in banner and ribbon name it the Person Dashboard. For everything else about person properties — how the subject is set up, how identity accuracy is judged, how narratives are tracked — see Tracking Individuals.

Best Practices

Read the explainer before you act on a number

Every card's info icon defines what is counted and what is excluded. Share of Voice deliberately ignores brand-identity and comparison prompts; message drift is better when it's lower. Acting on a metric you've misread costs more than the thirty seconds of hovering.

Set up a project before you trust the deltas

Without an active project, prompts only run when you generate a report, so the 7-day delta is comparing two report dates rather than two weeks. The dashboard says so, but the shortcut is: set up a project and let daily tracking fill in the line.

Annotate the same day you ship

The drill-down charts are only as good as the events on them. Add the annotation when you publish, not when you notice the metric moved.

Go to L3 when a number surprises you

A drop that looks alarming at L1 is often one prompt behaving differently, and you can see that in ninety seconds by opening the answers for the day it happened. Reading the actual response is faster than theorizing about the chart.

Work the opportunity modules in order

SEO opportunities, site readiness and earned media are three destinations for the same question — what should we do next. Gaps where a competitor already ranks are usually the cheapest win, site-audit findings the most mechanical, and earned-media pitching the slowest but most durable.

Related

• AI Visibility Reports — Baseline the property and populate every module
• Projects — Turn on daily tracking and set the goals that drive the goal-impact card
• Annotations — Put your actions on the drill-down charts
• Citation Intelligence — The raw-data level under the citation and earned-media modules
• Citation Optimizer — Where the SEO opportunity rows hand off
• Competitors · Locations · Key Messages — Configure the sets these modules measure
• Tracking Individuals — The person variant in full