Skip to content
Serpify
AI Visibility

See how AI assistants answer your buyers’ questions, and whether your brand appears.

Add the questions your buyers ask. Every week Serpify collects several answers to each one from ChatGPT, Gemini and Google AI Mode, then shows how often they mention your brand, cite your site and name your competitors, and which sources they draw on.

Start Free See How It Works
  • ChatGPT, Gemini, Google AI Mode
  • Several answers per question
  • Mention and citation rates
  • 95% confidence ranges
  • Sources and gaps

Placeholder illustration of the AI Visibility page with the AI Visibility Score, mention and citation rates with their ranges, share of voice and the weekly mention rate per engine, showing sample data only.

How it works

From a buyer’s question to your next step.

AI answers change from one ask to the next, so Serpify samples them, measures what they say about you, and points to what shapes them.

  1. 01

    Ask

    Add the questions your buyers ask AI assistants before they choose.

  2. 02

    Measure

    Several answers per question and engine show how often you are mentioned and cited.

  3. 03

    Compare

    See which of your competitors the answers name, and your share of the mentions.

  4. 04

    Understand sources

    See which pages the answers cite, and where they cite competitors but not you.

  5. 05

    Act

    Work through a weekly action plan of fixes, pages to write and sites to approach.

Your questions, in your market.

Questions belong to a project and are asked in its country and language. Add your own, or start from suggestions.

Paste or pick
Add up to 200 questions at once, or pick from suggestions built from your tracked keywords with fixed question templates. Suggestions are free and are not written by an AI model.
Organise
Label questions by topic, choose the engines for each, and pause the ones you don’t need.
Top questions, daily
Every question is asked weekly. Star your most important ones and they are also asked daily on ChatGPT, Gemini, Google AI Mode and Bing, up to your plan’s daily allowance.

Placeholder illustration of the Add prompts panel with pasted questions, a topic, engines and suggestions from tracked keywords, showing sample data only.

Engines

Which answers Serpify reads.

Every engine is read through DataForSEO, and each answer is stored with its text and the sources it cites. Questions are asked in the project’s market: ChatGPT, Gemini and Bing at country level, Google AI Mode at the project’s location.

AI assistants

  • ChatGPT DataForSEO’s capture of the answer the ChatGPT app gives, with web search on, and the sources it cites.
  • Gemini DataForSEO’s capture of the answer the Gemini app gives, and the sources it cites.
  • Google AI Mode Google’s AI Mode answer from DataForSEO’s search results API, and the pages it references.

Also monitored

  • Bing search All plans Bing’s answer blocks and its top 10 results, from DataForSEO’s search results API. This is a search results page, not an AI assistant’s answer, so it is labelled separately.

Mentions and citations

Named in the answer, or cited as a source.

Serpify records both for every answer it collects, for your brand and for each of your competitors, then rolls them up per question, per engine and per project.

Placeholder illustration of the prompt list with each engine's latest result as answers naming you out of answers sampled, citations and status filters, showing sample data only.

Each engine cell shows the latest run: answers naming you, out of the answers sampled.

Mention rate
The share of sampled answers that name your brand: one of its names, your domain or its distinctive root, as a whole word.
Citation rate
The share of sampled answers that cite a page on your domain among their sources.
By engine
Mention rate per engine with its range, and how many answers cited you.
AI share of voice
Your share of brand mentions across you and the competitors configured in the project.
AI Visibility Score
A 0–100 score from a fixed formula: mention rate (40%), citation rate (20%), share of voice (20%, once competitors are set up) and AI Readiness (20%). It is not produced by an AI model and does not predict rankings or traffic.

Confidence ranges

A rate, with the range it could really be.

Ask an assistant the same question twice and you can get two different answers. So Serpify never reports a single answer as the truth: it collects several and shows how sure the rate is.

Several answers per run
Each weekly run collects 3 answers per question on each engine, or 5 on Agency Plus.
Daily samples for top questions
Starred questions also get one answer a day on ChatGPT, Gemini, Google AI Mode and Bing.
95% confidence ranges
Mention and citation rates come with a 95% range (a Wilson score interval), overall and per engine. Fewer answers mean a wider range.

Placeholder illustration of mention rates overall and per engine, each with its 95% confidence range, showing sample data only.

Competitors

See who AI answers name instead of you.

AI Visibility uses the same competitor list as your Rank Tracker project. Every answer is checked for each competitor’s names and domain, as well as yours.

AI share of voice
Your share of brand mentions across you and the competitors configured in the project. Not a share of the whole market.
Competitor mentions and citations
Each answer lists the competitors it names, and its sources are labelled You or Competitor.
Questions you lose
Questions where answers rarely name you, with the competitors they name instead.

Placeholder illustration of AI share of voice for you and three competitors, and the questions you lose with the competitors named instead, showing sample data only.

Inside the answer

Read what the assistant actually said.

Open any question to read its latest answers on each engine, with your brand and your competitors highlighted, the sources each answer cites and the searches the engine ran to write it.

Placeholder illustration of a question's latest ChatGPT answer with brand names highlighted, competitors named, sources cited, the searches the engine ran and the run history, showing sample data only.

Full answer text
Every sampled answer from the latest run, with your brand and competitors marked.
Sources, labelled
The pages each answer cites, marked You or Competitor.
The engine’s own searches
The searches ChatGPT and Gemini ran to write the answer, where they report them.
Run history
How many answers named you on each earlier run of the question.

Sources

Find the pages that shape the answers.

Answers lean on a handful of pages. Serpify lists every domain cited in answers to your questions, so you can see which reviews, roundups, directories and forums the assistants read.

Cited domains and URLs
How many answers cite each domain, its most cited pages, and the questions and engines it appears for.
Source gaps
A third-party domain that appears in answers naming a competitor but not your brand. A place to look, not a promise of a placement.
Types and status
Each source is typed (review, listicle, directory, forum, social, publisher) from its domain and titles, with your placement status. Export the list to CSV.

Placeholder illustration of the Sources page with cited domains, their type, answers citing them, answers naming you, competitor-only answers and gap flags, showing sample data only.

Action plan

A short list of what to do next, every week.

Serpify turns what it found into a plain-language plan: a few fixes, pages to write, sites to approach and drafts to send. It is built by fixed rules from your own data, not written by an AI model.

Built from your data
Setup gaps, AI Readiness fixes, the questions you lose, source gaps and placement drafts waiting to be sent.
Ranked and ticked off
Items are ordered by priority. Mark them done or dismiss them, and the weekly rebuild keeps your choices.
Suggestions, not proof
The plan shows where to focus. Serpify does not measure whether a finished item changed your visibility.

Placeholder illustration of the weekly action plan with a crawler fix, a page to write, a site to pitch, a completed fix and a draft to send, showing sample data only.

AI Readiness

Make sure assistants can read and quote your pages.

A weekly check of your key pages, up to your plan’s page limit: your home page, the pages your keywords target, pages Google’s AI Overviews cite and your ranking pages.

Crawler access
Which AI crawlers your robots.txt allows or blocks, plus your llms.txt and sitemap.
Four kinds of checks
AI access, structured data, answer-ready content, and trust and technical, scored per page.
Fixes in order
One list across your pages, ranked by how many points each fix would add to the readiness score, with the score’s trend.

Placeholder illustration of the AI Readiness tab with the readiness score and trend, AI crawler access, category scores and fixes ranked by points gained, showing sample data only.

Content and placements

Turn lost questions into pages, and gaps into placements.

Start an answer article for a question you lose, or a directory listing or outreach pitch for a source gap, from an outline you fill in. You edit, approve and send them.

Answer articles
The answer up top, a comparison table, an FAQ and FAQ schema. Approve it, then export it as Markdown or HTML.
Listings and pitches
Directory suggestions start with the directories AI answers already cite, then a curated list for your kind of business.
You publish and send
Serpify never posts or sends anything for you. Track each placement from suggested to drafted, submitted and live.

Placeholder illustration of an answer article draft awaiting approval, with export buttons, beside placements at different stages, showing sample data only.

Alerts

Hear about a drop without checking every week.

AI alerts are checked after the answers for a run come in, and arrive as a digest.

  • Mentions fall

    When an engine’s weekly mention rate falls by the number of points you choose (20 by default) against the previous week.

  • Readiness slips

    When an AI search crawler is newly blocked in robots.txt, or the readiness score falls by your threshold.

  • Email, Slack, webhooks

    Choose email and Slack per rule. Webhooks receive every alert event.

  • Weekly by design

    Mention alerts compare weeks, so they follow the weekly runs rather than single answers.

Reports, API and MCP

Share AI visibility with clients and tools.

Branded PDF reports include an AI assistant visibility section with mention rate, citation rate, AI share of voice and the weekly mention rate per engine.

Scheduled and shared
Send reports weekly or monthly, download one any time, or share a read-only client link.
API
A read-only endpoint per project returns the AI Visibility Score, rates with their ranges, share of voice, top sources and the trend.
MCP
Ask for the same AI visibility data from an AI assistant through Serpify’s MCP server.
Explore Reports & Alerts

Placeholder illustration of the AI assistant visibility section of a branded PDF report, showing sample data only.

Methodology

How Serpify measures AI visibility.

AI answers vary from one ask to the next. Here is exactly how the numbers are made, so you know how far to trust them.

Sampled, not exhaustive
Serpify collects a sample of answers to the questions you add. The rates describe that sample, not every answer every person sees.
Collected through DataForSEO
Answers come from DataForSEO’s captures of the assistants and its search results API. Serpify does not call the assistants itself.
Several answers each time
3 answers per question on each engine in every weekly run, 5 on Agency Plus. Top questions get one more answer a day on ChatGPT, Gemini, Google AI Mode and Bing.
Weekly, in your market
Questions run weekly, early in the morning in your workspace’s time zone, in the project’s country and language. You can also ask a question on demand once a day.
Ranges, not false precision
Mention and citation rates show a 95% confidence range. Headline rates pool the latest answers for each question and engine, Bing search included.
Matched by rules
A brand counts as mentioned when one of its names appears as a whole word, and a site as cited when a source links to its domain. Set your brand names carefully: a name that is also a common word can match too often.

What the numbers don’t tell you

  • Where in an answer your brand appears, or its order in a list
  • Why an assistant chose one brand over another
  • Answers to questions you haven’t added
  • Whether a change you made caused a change in answers

Who it’s for

For teams who need to know how their brand shows up.

SEO agencies

Show each client how AI assistants answer their buyers’ questions, how often they are named against their competitors, and what you are doing about it, in a branded report.

For agencies

In-house SEO teams

Follow your brand’s mention and citation rates per engine, see which sources the answers lean on, and work through a weekly plan.

For in-house teams

Content teams

Find the questions you lose and the pages the assistants cite, then write answer pages for them.

For content teams

FAQ

AI Visibility, answered.

What does AI Visibility measure?
For each question you add, Serpify collects several answers per engine and records whether each one names your brand, cites a page on your site, names your competitors and which sources it cites. From those it reports mention rate, citation rate, AI share of voice, the most cited sources and an AI Visibility Score.
Which AI assistants does Serpify check?
ChatGPT, Gemini and Google AI Mode, plus Bing search, which is read from its answer blocks and top results and labelled separately because it is a search results page rather than an AI assistant’s answer.
How are the questions created?
You add them: paste up to 200 at a time, or pick from suggestions built from your tracked keywords using fixed question templates. Suggestions are not written by an AI model. Each question can have a topic label, its own engines, and can be paused.
How often are questions asked?
Every active question is asked weekly, with 3 answers per engine (5 on Agency Plus). Questions you star as top questions are also asked daily on ChatGPT, Gemini, Google AI Mode and Bing, one answer each, up to your plan’s daily allowance. You can also ask a question on demand once a day.
Why does Serpify show ranges instead of a single number?
AI answers vary between asks, so a rate from a sample of answers is an estimate. Serpify shows a 95% confidence range (a Wilson score interval) with mention and citation rates, overall and per engine.
What is AI share of voice?
Your share of brand mentions across you and the competitors configured in the project, in the latest answers. It is not a share of the whole market.
What is the AI Visibility Score?
A 0–100 score from a fixed formula that blends mention rate (40%), citation rate (20%), share of voice (20%, once competitors are set up) and AI Readiness (20%). It is not produced by an AI model and does not predict rankings or traffic.
Where does the data come from?
Through DataForSEO: its captures of the ChatGPT and Gemini apps, and its search results API for Google AI Mode and Bing. Serpify does not call the assistants itself.
Does Serpify tell me what to do?
Yes, as a weekly action plan built by fixed rules from your setup, AI Readiness fixes, the questions you lose, source gaps and placement drafts. It shows where to focus; Serpify does not measure whether a finished item changed your visibility.
Can Serpify write content for me?
Serpify gives you outlines for answer articles, directory listings and outreach pitches to fill in. You edit and approve them, export articles as Markdown or HTML, and publish and send them yourself.
How quickly do results update?
New questions get their first answers within minutes. After that, results update with each weekly run and, for top questions, each daily sample. AI alerts compare one week with the previous one.

Find out what AI assistants say when your buyers ask.

Add your first questions and see answers from ChatGPT, Gemini and Google AI Mode within minutes.

Start Free See How It Works