Weekly Benchmark · example
A dental practice in a mid-sized US city
Week 3 of 4 in this billing cycle · September 2026 · business anonymised
How this was measured
The same 50 questions its customers ask, once each on five AI tools, every week.
Every rate shown with its observed range across recent weeks.
Named, by AI tool
This week · observed range · direction
| AI tool | Named this week | Rate | Observed range | Direction |
|---|---|---|---|---|
| Perplexity | 23 of 50 | 46% | 33 to 60% | up on last week |
| Google AI OverviewsAI answer box present on 38 of 50 questions · rate over those · 0 collection failures, counted apart from zero | 6 of 38 | 16% | 6 to 24% | level with last week |
| ChatGPT | 5 of 50 | 10% | 4 to 21% | down on last week |
| Gemini | 3 of 50 | 6% | 2 to 16% | level with last week |
| Claude | 0 of 50 | 0% | 0 to 7% | level with last week |
A tool that gives no answers is left out rather than counted as zero. Ranges are the lowest and highest weekly values observed over the last 3 Benchmarks.
Position in this market
#4 of 12 practices AI tools name for these questions. On Perplexity:
| The market leader | 32 of 50 | 50 to 76% |
| Second practice | 26 of 50 | 39 to 65% |
| This practice | 23 of 50 | 33 to 60% |
| Median named practice | 6 of 50 | 6 to 24% |
The two practices above this one sit within a few points of each other: a cluster, not a fixed order.
What moved since last week
- New this week1
- Improved1
- Worsened0
- Unchanged3
- Resolved1
A finding keeps its identity week to week; a change of wording never counts as a new one.
Findings, highest value first
- 1Get citedUnchanged
ChatGPT reads Yelp and Healthgrades for dentist questions in your city. The two practices it names most appear on both. You appear on neither.
- yelp.com · cited in 31 of 50 answersThe market leader, Yelp profile, 312 reviews
- healthgrades.com · cited in 18 of 50 answersThe market leader, Healthgrades profile
Why this may matter: Consistent, complete business information across trusted local profiles is a standard local SEO practice and gives answer systems clearer source material to compare.
Suggested test: Complete both profiles, keep the business facts consistent, and recheck the same ChatGPT questions.
Evidence strength: Strong evidence · Measure next: Whether ChatGPT cites either profile and names the practice more often on the same controlled questions.
This is a well supported opportunity to test, not a promise that ChatGPT changes its answers.
Why this order: commercial importance high · competitive gap high · evidence strong · in your hands partly · 2 of 5 tools · confidence high
- 2Show upNew this week
You are missing from all 6 emergency questions. Every business named on them shows 24 hour service on its website. Yours does not mention hours.
- the leader's own site · cited in 6 of 6 emergency answersEmergency dentist: same-day care, open 24 hours
- the second practice's site · cited in 4 of 6 emergency answersAfter-hours and emergency care
Why this may matter: A focused, crawlable service page can make relevance, availability and location clearer to both search engines and answer systems.
Suggested test: Publish one clear emergency care page, then recheck the same six questions on the tools that named competitors.
Evidence strength: Strong evidence · Measure next: Whether the page is cited and whether the practice is named on any of the same six emergency questions.
The observed gap and SEO rationale make this test worthwhile, but no ranking or recommendation change is promised.
Why this order: commercial importance high · competitive gap widest · evidence strong · in your hands · 5 of 5 tools · confidence high
- 3Clean itImproved
Perplexity cited your competitor's reviews 31 times and yours once. Their review page is public and indexed. Yours sits behind a login.
- the leader's own site · cited in 31 of 50 answersPatient reviews, public and indexed
- your site · cited in 1 of 50 answersReviews, behind a login
Why this may matter: Content hidden behind a login cannot be crawled like a public page. Making useful review evidence accessible is consistent with basic technical SEO practice.
Suggested test: Publish an indexable review page, confirm it can be crawled, and recheck the same Perplexity questions.
Evidence strength: Moderate evidence · Measure next: Whether Perplexity cites the page and whether the citation gap changes on the same controlled questions.
Accessibility removes a clear technical barrier. It does not promise a citation or a recommendation.
Why this order: commercial importance medium · competitive gap high · evidence strong · in your hands · 1 of 5 tools · confidence medium
Order is Kensume's priority policy: what buyers ask about most, where the gap is widest, how firm the evidence is, what you can change yourself, how many AI tools it touches, and how confident we are. It is not a measured ranking lift and not a promised gain.
Observed evidence
A competitor has an FAQ answering this exact question. The suggested next step is a controlled test, not a promised outcome.
Expert rationale
The opportunity is checked against established SEO and content practices, then measured again with the same questions to see what changed.
What to measure next
The next Benchmark compares the same tool, questions, citations and named businesses. If nothing changes, the report says so.