The Case Study
I Tested 12 Buyer-Intent Queries
On My Own Site
Most AI visibility consultants ask you to trust their methodology. I publish the test on my own site. Honest before-data, dated and updated.
Day 0 baseline: 2026-05-15 for Google AI Overviews (captured via verifiable web search). ChatGPT, Perplexity, and Claude baselines pending direct capture during the public Day 0 run. Reruns of all four tools at Day 30, 60, 90, and 180.
Day 0 headline result:lesli.com is invisible on 10 of 12 buyer-intent queries in Google AI Overviews, cited with a stale “success coach” description on 1 query (entity disambiguation failure), and named as a passing reference on 1 query (with a trademark naming-conflict risk on “AI Visibility Stack”). This is the honest before-data the methodology is supposed to fix.
Why This Page Exists
Every AI visibility consultant claims to get clients cited by ChatGPT, Perplexity, and Google AI Overviews. Almost none of them publish the test on their own site.
The reason is straightforward. The before-data is usually embarrassing. The consultant's own site is invisible on the buyer-intent queries that should surface it. So the test stays private, the methodology stays unverifiable, and the buyer has to trust a logo wall.
This page reverses the pattern. I run 12 buyer-intent queries against four AI tools, dated, and publish the raw results. If the methodology works on my own site over 90 to 180 days, the same methodology works on yours. If it does not, you will see that too.
The page is a living dataset. The Day 0 baseline is the truth as it stands today. The deltas at Day 30, 60, 90, and 180 are the proof.
How
The Methodology
1. Pick 12 buyer-intent queries
Each query mirrors a real question a Series A SaaS founder would ask an AI tool about an AI visibility consultant. Mix of high-commercial intent, named-entity checks, methodology lookups, competitor displacement, and definitional questions.
2. Run each query against four AI tools
ChatGPT (GPT-4 default), Perplexity (Sonar default), Claude (latest available), and Google AI Overviews (logged-out, Canada-IP, English). Capture verbatim responses, cited URLs, and competitors mentioned.
3. Score each result
Cited (lesli.com or Lesli Rose named, with link or unambiguous attribution). Partial (named without link, or with inaccurate description). Not cited (no mention). Baseline pending (test not yet executed for Day 0).
4. Publish the raw data
Every query, every tool, every score, dated. No editing. No selective reporting. The dataset is licensed CC-BY-4.0 so anyone can replicate the methodology.
5. Rerun at Day 30, 60, 90, and 180
Same queries. Same tools. Same scoring. The deltas reveal whether the methodology is moving the needle. The page is the longest-running public dataset of any AI visibility consultant in Canada.
The methodology behind which queries to pick and how to engineer pages for citation is at the named-query methodology page.
The Data
The 12 Queries
Day 0 baseline established 2026-05-15for Google AI Overviews. ChatGPT, Perplexity, and Claude statuses currently show “Baseline pending” and will be captured during the public Day 0 run. Each result includes the competing sources cited so you can verify the test yourself.
“Best AI visibility consultant for Series A SaaS in Canada”
Intent: high-commercial buyer intent, geo + stage qualifier
GrowSERP, Stridec, Ranking Digitally cited. lesli.com absent.
“Who is Lesli Rose?”
Intent: named-entity check, entity disambiguation
lesli.com cited, but described as 'success coach and David Bayer Certified Transformational Mindset Facilitator' (stale positioning). Entity disambiguation failure.
“AI visibility consultant who works with VC-backed startups”
Intent: high-commercial, vendor-aware, audience qualifier
Hotwire and Sparkpr cited. lesli.com absent.
“How to get a SaaS site cited by Perplexity”
Intent: how-to, solution-aware, methodology-shopping
authoritytech.io, ferventers, ailabsaudit, averi.ai, discoveredlabs cited. lesli.com absent.
“Canadian AEO consultant for funded startups”
Intent: high-commercial, geo + stage + category
Jordan Stevens, Think with AEO, ThatWare cited. lesli.com absent.
“AI Visibility Stack methodology”
Intent: named-methodology check, branded category
lesli.com /insights/* cited as passing reference. Primary citation goes to Gabriela Marco's SSRN paper, which uses the same trademark for a different (3-layer agent-discovery) framework. Naming-conflict risk.
“Series A SaaS marketing consultant Canada”
Intent: high-commercial, stage + geo + role
7 Eagles, Roketto, Bay Leaf Digital, The Marketing Agency cited. lesli.com absent.
“Alternatives to RankScience for AI visibility”
Intent: competitor-displacement, vendor-aware
Query collapses RankScience with Rankscale (tool, not agency). Peec AI, Otterly.ai, SE Ranking, Scrunch AI cited. lesli.com absent.
“AI visibility for govtech Series A”
Intent: vertical + stage, white-space query
Concurate's 9 GEO Agencies for GovTech SaaS roundup cited. lesli.com absent.
“What is an AI visibility consultant”
Intent: definitional, top-of-funnel, category-aware
Eric Schwartzman, Jason Khoo, AdsX, allaboutai cited. lesli.com absent.
“Best SEO consultant for B2B SaaS post-Series-A”
Intent: high-commercial, stage qualifier, role displacement
RevenueZen, Grow and Convert, Directive Consulting, Skale, Stratabeat cited. lesli.com absent.
“Should I hire an AI visibility consultant after my Series A?”
Intent: decision-stage, founder-direct question
Eric Schwartzman, AdsX, Jason Khoo, Ranknizer, Cassie Clark cited. lesli.com absent.
What This Will Prove (or Disprove)
One of three things will happen over the next 180 days.
Option A: Citations grow across most tools and most queries. The methodology works. The page becomes the strongest single piece of proof in my portfolio.
Option B: Citations grow on some queries and not others. The methodology works for certain query types and not others. The page documents which patterns work and which do not. That is also useful.
Option C: Citations do not grow. The methodology does not work, or the test was constructed wrong, or lesli.com is too new to compete for these queries. In that case, I update the methodology publicly, repeat the test, and publish the revised results.
None of these options are bad outcomes. Option A is the marketing story. Option B is the operational nuance. Option C is the credibility move that most consultants will never make.
How to Replicate This on Your Own Site
The dataset is licensed CC-BY-4.0. Replicate it on your own Series A SaaS site as follows.
- Pick 12 buyer-intent queries that mirror what a buyer in your category would ask an AI tool about your kind of company. Mix high-commercial, named-entity, methodology, competitor-displacement, and definitional queries.
- Run each query against ChatGPT, Perplexity, Claude, and Google AI Overviews. Logged out, fresh session, no personalization.
- Capture the verbatim response, the cited URLs, and the competitors mentioned. Save as text or markdown.
- Score each result Cited, Partial, or Not cited.
- Publish the raw data on a public page on your site. Date it. License it CC-BY-4.0.
- Rerun at Day 30, 60, 90, and 180. Update the same page. Never delete the historical data.
If you want help interpreting the data or engineering pages to move the citations, that is what I do. Book a strategy call or see Series A pricing.
Want This Run on
Your Site?
The $297 Visibility Action Plan includes a 12-query baseline against ChatGPT, Perplexity, and Google AI Overviews for your business, plus three competitor comparisons. Book a strategy call first if you want me to interpret the data with you.