AnnouncementThe Infinity plan now tracks 30 150 prompts for your brand ๐ŸŽ‰

Features / Prompt Tracking

AI Prompt Tracking That Shows When AI Recommends Your Brand

Track the questions your buyers ask ChatGPT, Gemini, Claude, Perplexity and Google AI Overviews. See which prompts name your brand, which cite your pages, and which ones a competitor gets instead.

5.0 on PeerPushJoined by 400+ teams
Named and cited
23
57% of 40 runs
Named, not cited
2
5% of 40 runs
Cited, not named
5
13% of 40 runs
Neither
10
25% of 40 runs
Visibility mix
8 prompts ร— 5 engines = 40 runs
Mention rate62.5%
Citation rate70.0%
Named and citedNamed, not citedCited, not namedNeither
Changes since Nov 5
2 Lost visibilityView
3 Gained visibilityView
3 New promptsโœ“ 3 visibleView
Category: AllVisibility: AllPrompt Explorer 8
Prompt ResultsSearch Prompts8 of 8
PromptLatest ยท Nov 12, 9:40 AMSentimentCompetitorsSince last run
Reading the engine glyph
Named and cited
Named, not cited
Cited, not named
Neither

Left half is the name in the answer text. Right half is a brand URL among the sources.

Trusted by
Grey AestheticsLuminarymyIDFi AI
01
Define
The basics

What is AI prompt tracking?

AI prompt tracking is running the same buyer questions through AI assistants every week and recording what comes back. Is your brand named? Is one of your pages cited as a source? Which competitors show up, and did any of that change since last week?

It does the job rank tracking does for Google, with one big difference. An AI answer has no positions 1 to 10. It names two or three brands and moves on, so the useful measure is how often youโ€™re one of them.

One check in ChatGPT tells you very little, because answers vary from run to run and from engine to engine. Tracking the same set of prompts over time is what turns those answers into a number you can report and improve.

Keyword ยท Google

project management software

Ten blue links. You rank #4 and still get some clicks.
Prompt ยท ChatGPT

โ€œWhatโ€™s the best project management tool for a 12-person agency that bills clients hourly?โ€

TasklineYourbrandOrbitlyThree names. Youโ€™re either one of them or youโ€™re invisible.
5AI engines tracked: ChatGPT, Google AI Overviews, Perplexity, Gemini and Claude
4answer states per run: named and cited, named, cited, or neither
Weeklyre-runs, plus a new run any time you want one
8prompt types, from branded and comparison to local and informational
02
Answers
Prompt detail

Read the full answer from every AI engine

Click any prompt and the full answers open, one tab per engine. Your brand and your competitors are highlighted, every source is listed, and the history shows how the answer has changed run by run.

03
Read
Answer states

Named, cited, both or neither

Every answer lands in one of four states. Each one points to a different fix, so you always know what to work on next.

Named and cited23 of 40 runs

Named and cited

The engine recommends you and links to one of your pages as the source. This is the result you want on every prompt.

What to doKeep the cited page current and watch the next run for drops.
Named, not cited2 of 40 runs

Named, not cited

The engine knows your brand but takes its facts from someone else, often a review site or a competitor's blog.

What to doPublish a page that answers the question directly, so the engine has something of yours to cite.
Cited, not named5 of 40 runs

Cited, not named

The engine uses your page to build its answer but never says your name. You did the work and someone else gets the credit.

What to doPut the brand in that page's heading, first paragraph and structured data.
Neither10 of 40 runs

Neither

You're missing from the answer entirely. The engine named other brands and used other sources.

What to doTreat it as a content gap. The Blog Agent suggests a post written for this prompt.
04
Choose
Your prompt set

Track the prompts that matter to your business

  • A starting set written from your site

    Botric reads your site and brand profile, then writes the questions buyers ask across branded, commercial, comparison, informational, transactional and local intents.

  • Lock the prompts you canโ€™t lose

    Botric refreshes its suggestions on each run so the set follows what buyers ask. Lock any prompt and it stays put, run after run.

  • Add fixed prompts of your own

    Paid plans include five fixed prompts per site. Write them yourself and Botric never swaps them out.

  • Test before you track

    Try new prompts in Prompt Research first, read the answers, and only import the ones worth a tracked slot.

05
Compare
Run history

See what changed since the last run

Every run is compared with the one before it. You see which prompts lost visibility, which gained it, and which are new, without scrolling through 125 answers to work it out.

  • Catch drops early

    A prompt that slips from named to missing shows up the same week, while thereโ€™s still time to fix the page behind it.

  • Prove the work landed

    Publish a page, then watch the next runs. The history shows when an engine started naming or citing you.

06
Setup
How it works

How does AI prompt tracking work in Botric?

Add your website and Botric runs the whole analysis. It reads your site, finds competitors, writes buyer questions and runs every one on your AI engines.

No code or tag to install. Results in minutes.
OverviewRecommendationsActionsRunningโ€ฆ

Analyzing your AI visibility

Running prompts across LLMs

Querying your plan's AI engines

45%
Reading websiteFinding rivalsTopicsBuyer questionsLLM promptsCitationsCompetitorsSentimentCompetitive analysisReport

You can leave this page โ€” we keep working

08
Pricing
Plans

Prompt tracking on every plan

Plans differ in how many prompts you track and how many engines they run on. Every paid plan starts with a 7-day free trial.

For solo teams
Bit

A minimal starting point for testing AI visibility and lightweight automation.

$15/ month
Engines

ChatGPT, Google AI Overview

15 tracked prompts
Prompt Research
1,500 credits per cycle
1 workspace ยท 1 site ยท 1 admin seat
Up to 6 blogs per month
Not included: MCP & API access
Scale-ready
Infinity

Teams running multiple properties and deeper AI growth workflows.

$149/ month
Engines

ChatGPT, Google AI Overview, Perplexity, Claude, Gemini

150 tracked prompts
Prompt Research
15,000 credits per cycle
5 workspaces ยท 5 sites ยท 5 admin seats
Unlimited blogs
MCP & API access
09
Proof
Case study ยท MotorEnvy

AI answers naming MotorEnvy up 80% in three months.

Botric mapped the questions luxury car buyers ask AI, measured how often eight AI platforms named MotorEnvy, and published pages for the prompts where competitors were named instead.

+80%
AI answers naming MotorEnvy
8
AI platforms naming the brand
61
Keywords at #1 on Google
Read the case study โ†’
10
Questions
Questions

Frequently asked questions

What people ask before they start tracking prompts.

Q1What is AI prompt tracking?

AI prompt tracking means running a fixed set of buyer questions through AI assistants on a schedule and recording what comes back: whether your brand is named, whether your pages are cited, which competitors appear, and how each answer changes from one run to the next. It is the AI search version of rank tracking.

Q2How is a prompt different from a keyword?

A keyword is a short search term, and Google answers it with a ranked list of links. A prompt is a full question, often with context like team size, budget or location, and the AI answers it with a few named recommendations. There is no position 4 in an AI answer: you are either one of the brands it names or you are not, which is why prompts are tracked by mention and citation rather than by rank.

Q3Which AI engines does Botric track prompts on?

Botric currently tracks ChatGPT, Google AI Overviews, Perplexity, Gemini and Claude. Coverage depends on the plan: Free covers ChatGPT, Bit adds Google AI Overviews, Byte adds Perplexity, and Infinity covers all five.

Q4How many prompts can I track?

Free tracks 5 prompts, Bit 15, Byte 25, and Infinity 150 across up to five sites. On paid plans each site includes five fixed prompts that you write yourself and that Botric never swaps out.

Q5Where do the prompts come from?

When you add your site, Botric reads it, finds your competitors, groups your topics and writes the questions a buyer would ask, across branded, commercial, comparison, informational, transactional and local intents. You can lock any prompt so it stays in your set, add your own, and test new ones in Prompt Research before they take up a slot.

Q6How often are prompts re-run?

Tracked prompts are re-run every week, and you can start a new run from the dashboard at any time. Each run is compared with the previous one, so you can see which prompts gained visibility, which lost it, and which are new.

Q7What does "cited, not named" mean?

It means an AI engine used one of your pages as a source for its answer but did not say your brand name in the text. It usually happens when the page explains a topic well but never ties it back to the brand. Making the brand explicit in that page's heading, opening paragraph and structured data is the usual fix.

Q8What should I do with a prompt that never mentions my brand?

Treat it as a content gap. Open the prompt to see which brands and sources the engines use instead, then check Citations for pages that cite competitors but not you. The Blog Agent also suggests topics written for the prompts you track, so each new post targets a prompt you're missing from.

11
Start
Free AI visibility audit

Find out which prompts mention you today.

Run a free audit to see how ready your site is for AI search, then start a trial to track your prompts every week.

โ— Free ยท results in 30 seconds ยท no signup required