company logo

Help center

Go to Rank Prompt
Powered by
All collectionsReportsUnderstanding report results

Understanding report results

How to read a finished report: the Visibility score and citation share, the five tabs, each answer in full, and what to do with what you find.

A report page has three parts: the header with your brand and filters, the score card with two charts, and five tabs. Everything on the page follows the category and AI platform filters at the top.


The top of the page

  • Your brand and its aliases: the other names or spellings AI answers use for your brand. An answer that names one counts as naming you. Click Add alias to add one; when Rank Prompt spots a likely alias in the answers, it suggests it for you to accept or dismiss

  • Filters: narrow the page to the prompt categories and AI platforms you pick, one or several

  • View as a competitor: see the whole report as if a competitor were the focus brand. Nothing is saved; it is only your view

The actions sit in the header: Schedule, Share, Export PDF, and under More, Export JSON, Export CSV and Campaign.


The score card

Visibility score

Out of every AI answer collected, the share that named your brand. Under the gauge you see the count, for example "9 of 30 answers". Each prompt on each AI platform is one answer.

On the AI visibility page, a score also gets a tier:

Tier

Score

What it means

Emerging

Under 25%

There is a wide opening to get AI platforms to name you

Moderate

25% to 49%

You show up for some prompts, with real room to grow

Strong

50% and up

AI platforms name and cite you regularly when people ask about your category

Low numbers are normal: most brands start in single digits. Hover the gauge and, once enough brands have run a first report, it also says the score that half of brands stay at or under on their first report, and the one three in four stay at or under. Rank Prompt works both out once a day from every brand's first report, and never shows how many brands it counts.

Citation share

The second gauge, Citation, shows how often AI cited your own site as a source, out of every source it cited in the report. It appears when your brand has a website and the answers cite sources.

The charts

  • Visibility by platform: your score on each AI platform in the report

  • Visibility by category: your score in each prompt category

Different AI platforms read different sources, so the same prompt can name you on one and miss you on another.


Prompts

The prompts the report asked and their answers. The header counts how many were Named and Not named.

Column

What it shows

Prompt

The prompt that was asked

Status

Named or Not named

Best rank

Where your brand sat in the answer (1 is first)

Brands named

The other brands in the answer

Sources

The pages the answer cited

Date

When it was answered

Use All prompts, Named only or Not named only, and search by prompt. Click a row to read the answer in full.

Two views sit above the table: Query map groups your prompts into clusters, and Query graph shows what each AI platform searched and the related questions it surfaced.

Reading an answer

The answer opens with your brand marked and shows:

  • Whether it named you, and at which place

  • Brands named: every brand, product or entity in the answer, sorted into your brand, competitors, publishers or other

  • The sources it cited (Show sources)

  • More from the AI platform: images, follow-up questions, searches and other blocks it used

  • Previous answer and Next answer to move through the report


How to fix your results

Click Show me what to fix and Rank Prompt reads the whole report and returns the fixes as tasks: quick wins, the topics to cover, and what the pages AI cites instead of you do that yours do not.

  • Tasks from this report arrive as suggestions: accept the ones you want onto your board and dismiss the rest, or Open in Tasks

  • Source insights analyzes the pages AI cites instead of your brand, with What to copy: from each

  • Regenerate writes a new set

The first run on a brand is free; after that it costs 1 credit, and so does Regenerate. Only owners and editors of the brand can run it.


Competitors

The Share of voice ranking for this report: your brand and every business the answers named.

  • Share of voice: of all the times AI named you or a competitor, the share that was you

  • Answers: the answers that name it

  • Won without you: the answers that name it and leave you out

Competitor types picks who is counted, one type or several: Direct competitor, Indirect competitor, AI platform, Publisher or Other. It starts on Direct and indirect.

Click a competitor to see its share against yours, the Platforms where it beats you, the prompts it wins without you, and the domains that cite it and not you. From there you can View as this competitor or Change category if it is sorted wrong. Co-occurrence shows which brands AI names together.


Sources

The sites and pages the answers cited, the same views as Sources > Citations, limited to this report:

  • Top cited domains, and each domain's Citation share

  • The kinds of source (your site, competitors, reviews and directories, news, forums, Reddit and more); click one to filter

  • Cited pages, with each page's citations, position, the AI platforms that cited it and your outreach on it

Every cited page gets a citation card after each report, free on every plan: the page's image and site name, and Author and Published columns. For a LinkedIn or Reddit page, Show post opens the post inside the app.

To pitch these sites, click Open in Backlink outreach, or select pages and click Create campaign (Pro and up).


By platform

Every prompt against every AI platform in the report: Yes with your rank when the answer named you, No when it did not, and N/A when there is no answer from that AI platform. Turn on Only named to keep the prompts that name you somewhere, and read the Keywords column for the Google keyword each prompt maps to.


While a report runs, and when it stops

  • While it runs, the page fills in as the answers arrive, with an early read of the score. The fixes, exports and share link open once the report finishes

  • If it did not finish, it keeps the answers it saved as partial results. They do not count in Visibility or its charts

  • If the analysis failed, Retry analysis runs it again for free


Taking action

If your score is low

  1. Start with Show me what to fix

  2. Read the answers that do not name you, and the brands and pages they name instead

  3. Write content for the prompts that miss you

  4. Earn mentions on the sites the answers cite

If your score is high

  1. Keep the pages AI cites up to date

  2. Add prompts for the categories you do not track yet

  3. Put the report on a schedule and watch for drops

  4. Watch the competitors that are rising

See Improve your results.


Exporting and sharing

  • Export PDF: the report as a document, on every plan, branded with your white label settings from the Agency plan

  • Export CSV: the answers as a spreadsheet

  • Export JSON: the raw results for your own tools

  • Share: a read-only link (see Shares)

For a branded document you publish for a client every month, use Visibility > Client reports.


Questions?

Email [email protected] and I'll help.

Did this answer your question?
😞
😐
😁