AI mention trackers
Claude mention tracker over time
Claude mention tracker that shows when Anthropic's Claude names your brand, what it cites, and how it describes you over time.
- Free first report
- No credit card
- 1 credit covers 3 engines
Trusted by teams and agencies winning AI search
Rank Prompt is a Claude mention tracker that asks Claude your buyers' questions on a schedule, with web search on. Each run logs whether Claude named you, whether it cited your own pages, and who it recommended in your place.
Know when Claude names you
Each Claude answer is split into named and cited, named only, or not named.
-
best running shoes for beginners
Named · cited
Claude -
best shoes for a first marathon
Named, not cited
Claude -
cheapest good running shoes
Not named
Claude
- 1
Nike 17 / 30 - 2
Brooks 17 / 30 - 3
HOKA 16 / 30 - 4
adidas 13 / 30 - 5
ASICS 11 / 30 - 6
New Balance 4 / 30
Claude mentions, with sources
Reports run Claude with web search, so each mention comes with the pages it read.
- 01
Add your company
Rank Prompt reads your site and suggests categories and prompts; the first report shows who Claude names in your place.
- 02
Pick the prompts
Include brand questions, where Claude describes your company in its own words.
- 03
Claude runs by default
With ChatGPT and Google AI Mode (AI Overviews), for 1 credit per prompt, on your schedule.
- 04
Read the answer
Whether Claude named you, which listings it cited, and how it describes you.
Claude answer Named #2 Youbest running shoes for beginners
These come up most often:
- Brooks: a stable, forgiving ride that holds up over the miles
runnersworld.com - adidas: responsive cushioning, a roomy fit and good value
reddit.com - ASICS: comfortable, with extra support if you need it
youtube.com
Sources (3)
- 1
runnersworld.com - 2
reddit.com - 3
youtube.com
In every answer
- Mentions on Claude How often Claude named you, per prompt
- Your pages cited How often Claude used your own site as a source
- Listings it read The Wikipedia, Yelp and directory pages behind each answer
- Brands it prefers Who Claude recommends instead, and who is new in recent runs
When Claude started or stopped naming you
Two runs side by side, with the exact prompts that changed hands.
Run of 1 September 43 13 of 30 answers
Run of 4 August 33 10 of 30 answers
Newly won
Newly lost
How Claude cites
Averages across brands tracked in Rank Prompt over the past year. The sample is smaller than for ChatGPT, so treat it as a rough guide.
- 12 sources per answer
- 96% answers cite at least one source
- 20% of its sources also cited by Perplexity
What Claude knows without searching
Reports run Claude with web search. The Brand Audit also asks it without search, so you see what it already believes about your brand and where that is out of date.
What it means for you
- What Claude knows about you See Claude's description of your brand from memory, with its sentiment.
- Newer than its training? If Claude does not know you yet, work the search engines first. Training catches up in months.
- Search-backed in reports In reports Claude searches the web, and Fan-outs list those searches.
- Follow the trend Weekly runs show the trend, and the weekly email digest flags when your visibility moves.
Measure it, then close the gaps
Each answer shows what the engine cited. Use it to win back the prompts you lost.
Track mentions engine by engine
Every plan includes Claude and the other 5 engines, so you can compare them in one report.
How do I track brand mentions in Claude?
Claude runs in every report by default. Add the questions your buyers ask and a schedule; each run logs whether Claude named you, whether it cited your pages, and the rivals and sources in the answer.
Can I see what Claude says about my brand?
Yes. The Brand Audit (Starter plan, from 10 credits) asks Claude about your brand from memory and shows its description and sentiment, next to the models that search the web.
How much does Claude mention tracking cost?
No extra cost on the defaults: Claude, ChatGPT and Google AI Mode (AI Overviews) share 1 credit per prompt per run. Starter at $49/mo covers about 150 prompt runs on them.
How many sources does Claude cite per answer?
About 12, since Rank Prompt runs Claude with web search, and 96% of answers cite something. Claude leans on Wikipedia and business listings more than the other engines do. The sample is smaller than for ChatGPT, so treat it as a rough guide.
Free AI visibility report first · 7-day trial on any plan
Track your mentions in Claude
Claude runs in your first report for free. Add your brand to start.
- Starter plan/month
- $49
- Credits / month
- 150
- AI platforms tracked
- 6