Part A — Verdict
Overall AI presence
Partially recommendedTested 5 prompts across 10 AI answers. Your brand appeared in 4 answers.
Recommendation share 40% · estimated range 17–69% · small sample, directional signal
Broad category — product analytics
Mentioned in 4 of 10 broad product analytics answers.
Niche category
Not tested this run. Recommended as the next measurement.
Strategic summary
posthog.com appears in 4 of 10 tested product analytics answers. It is on the map but not a default recommendation — targeted category and comparison content can raise its share.
Part B — Evidence
The evidence behind the verdict
What we asked, what each AI answered, and who it named instead. Every number below traces to the 10 answers we generated for this report — the full answers are in the heatmap underneath.
Across 10 AI answers to 5 buyer-intent prompts, posthog.com was recommended in 4 of them.
That covers the broad category we tested — product analytics. Only this one broad category was measured this run; see "Scope of this test" below.
Recommendation share 40% · estimated 95% range 17–69% · small sample, directional signal
Prompts tested
5 prompts · 10 answersWe asked each of these on ChatGPT and Claude — the exact questions, shown in full below.
Recommended instead
14 namedIn the 6 answers where posthog.com was not recommended, these product analytics resources were named instead, ranked by how many cited each.
+8 more named
Sample response
Excerpt — full answers open in the heatmap below.Why it scored this way: The response does not include any mention of posthog.com, resulting in a score of 0 for overall quality. It provides a detailed list of other analytics tools but fails to address the target brand in any capacity.
posthog.com does not appear in this answer.
Scope of this test
Broad category — product analytics
Mentioned in 4 of 10 broad product analytics answers.
Niche category
Not tested this run. Recommended as the next measurement.
We tested the broad category — “product analytics”— because it is the most contested place to appear and the hardest first test to pass: the generic questions a buyer asks before they’ve named a specific tool. So the 40% above is a precise, limited finding about that broad category only. It does notprove the site is invisible in its own narrower niche — only the broad category was measured this run. A follow-up run should test prompts specific to the site’s real category.
Tested = measured on real AI answers this run. Not tested = no answer was generated, so no score is claimed.
All 10 answers are in the heatmap below — click any cell to read the full response.
Part C — Diagnosis
Diagnosis
Why AI assistants are underrepresenting your brand
Competitors are recommended in answers where your brand was absent. Assistants surfaced these named alternatives for questions where your brand did not appear.
Recommended instead of you in 6 answers: Mixpanel, Amplitude, Heap.
You perform well on some AI platforms but are largely invisible on others. Training data and knowledge cut-offs vary by model.
Claude: 60% vs ChatGPT: 20% — 40 point gap.
How each AI sees you.
Different platforms have different knowledge bases. Improving one often lifts the others.
ChatGPT
ChatGPT recommended Mixpanel and Amplitude, among others instead of you.
Claude
Claude recommends you in most tested answers.
Visibility Heatmap
How well does each AI platform know you across different prompt types?
| Prompt Type | ChatGPT | Claude |
|---|---|---|
Direct Recommendation What are the best product analytics options available right now? | 0 | 7 |
Problem Solution I need a product analytics solution. What do you recommend? | 0 | 9 |
Comparison Request Can you compare the top 5 product analytics options for me? | 0 | 0 |
Beginner Guidance I'm new to product analytics. Where should I start? | 0 | 0 |
Budget Constrained What's the best product analytics for someone on a budget? | 6 | 9 |
See exactly how your score was built
The per-prompt math behind your score — which prompts moved it, and by how much.
- Full per-prompt contribution breakdown
- Weekly remeasurement across both assistants
- CSV / PDF export of the full table
Flat monthly price. Cancel anytime. Your current report is saved.
Unlock the full competitive breakdown
Mention quality, ranking, share of voice, statistical confidence and competitive positioning — with 95% confidence intervals.
- Full 5-section competitive breakdown with confidence intervals
- Weekly remeasurement + proof of before/after share
- PDF / CSV export
Flat monthly price. Cancel anytime. Your current report is saved.
One fix, ranked by expected impact.
Built from the answers we captured — what assistants recommended instead, and the page format they cite when they do. Each fix names where it goes and why it earns the recommendation.
Publish a comparison page vs the named alternatives
Supporting5 minutesChatGPT recommended Mixpanel instead of posthog.com when asked "Can you compare the top 5 product analytics options for me?". Assistants pull the name they can most clearly match to that question — the fixes below give posthog.com that match for "product analytics".
Mixpanel appears here instead.
“Certainly! Here’s a comparison of five popular product analytics tools: **Mixpanel**, **Amplitude**, **Heap**, **Google Analytics**, and **Pendo**. Ea...”
Where: A new /compare page (e.g. yoursite.com/compare)
Sign in to track this fix
Every Monday: your highest-priority fixes, ranked — and proof of whether each one moved your recommendation share.
Get every fix written out — and watch each one move your share.
Every fix comes with paste-ready copy and exact placement. Then we re-run all your prompts across ChatGPT and Claude each week and show you which fixes actually moved your recommendation share — proof, not a dashboard.
Flat $99/month · cancel anytime · remeasured weekly.