Across 39real “hire an AI visibility consultant” prompts and 3,985 recorded citations (Otterly.AI, Canada, 2026-08-24–2026-08-30), hamitahm.com was cited 188 times: the single most-cited domain out of 1,046 domains that appeared anywhere in the panel. But it was named as the actual recommendation in only 1 of 39prompts. Both numbers are real. Neither one is the whole picture; see “What this isn’t claiming” below.
I ask clients to let me measure them honestly, including the parts that don’t flatter them. So I ran the same method on myself first. Every number below comes from a raw export you could recompute yourself; the source files are described in the methodology section.
01: What was actually measured
39real buyer-intent prompts about hiring an AI visibility consultant, an AEO agency, or a GEO specialist (“Who are the best AI visibility consultants in Canada?”, “Can you recommend an AI visibility consultant in Toronto?”, and 37 more like them) were run through ChatGPT, Perplexity, Google, Microsoft Copilot, for the Canada market, over the week of 2026-08-24 to 2026-08-30, using Otterly.AI, a competitive AI-search tracking tool. Every domain each engine cited as a source was logged, along with how many times: 3,985 total citations, across 1,046 distinct cited domains.
Otterly separately maintains a Brand Ranking leaderboard of about 10 named competitor brands for this panel, ranked by mentions and sentiment, generated from a filtered view (13 of the 39 prompts, its own “last 14 days” default). The domain-citation numbers in this post come from the full 39-prompt export; the Brand Ranking figures are labeled separately below and aren’t mixed with them.
This is a different instrument and a different dataset from the Bing Webmaster Tools / Google Search Console numbers behind the “AI engines already name Hami Tahm” case study and the HomeCalc AI search benchmark. Those measure a different site, a different category, or actual search-console traffic over months. This measures one competitive prompt panel, one category, one week, through one third-party crawler. None of these are added together anywhere on this site.
02: The domain ranking
Sorted by total citations (summed across the whole week) across all 1,046 domains that showed up anywhere in the 39-prompt panel. hamitahm.com is 1st, ahead of a Semrush agency directory, YouTube, Google’s own developer docs, LinkedIn, and every named competing agency’s own domain, including Longhouse, the agency Otterly’s separate Brand Ranking table (below) ranks 1st by mentions.
| # | Domain | Kind | Citations |
|---|---|---|---|
| 1 | hamitahm.com | HamiTahm (this site) | 188 |
| 2 | agencies.semrush.com | Directory (Semrush) | 131 |
| 3 | youtube.com | Video platform | 114 |
| 4 | developers.google.com | Google documentation | 78 |
| 5 | linkedin.com | Social platform | 68 |
| 6 | help.openai.com | OpenAI documentation | 59 |
| 7 | thatware.co | Competitor | 52 |
| 8 | longhouse.co | Competitor | 43 |
| 9 | semrush.com | Directory (Semrush) | 43 |
| 10 | lumirank.ca | Competitor | 41 |
Source: Otterly.AI, Canada, 2026-08-24–2026-08-30, 39 prompts, ChatGPT / Perplexity / Google / Microsoft Copilot.
The pages doing the work are the AI Visibility hub, the Canada and Toronto consultant pages, the audit page, and the pricing page, in that order; see the table in the next section. This rank is a citation count across one competitive panel, not a claim about traffic, revenue, or market share.
03: The honest gap: cited, but not recommended
Where the citations come from
Five pages account for 168 of hamitahm.com’s 188 citations.
| Page | Citations |
|---|---|
| /ai-visibility/ai-visibility-consultant-canada/ | 50 |
| /ai-visibility/ | 37 |
| /ai-visibility/ai-visibility-consultant-toronto/ | 35 |
| /ai-visibility/ai-visibility-audit/ | 29 |
| /pricing/ | 17 |
Where it shows up most
hamitahm.com was cited in 10 of the 39 tracked prompts. Its strongest questions are specific and commercial, not generic:
| Prompt | Citations |
|---|---|
| “Who offers a flat-fee AI visibility audit without a monthly retainer?” | 36 |
| “Who are the best AI visibility consultants in Canada?” | 32 |
| “Where can I get an AI visibility audit in Canada?” | 29 |
| “Which AI visibility consultants work with healthcare clinics in Canada?” | 27 |
| “Which consultants provide an AI visibility audit and prioritized action plan?” | 26 |
| “Can you recommend an AI visibility consultant in Toronto?” The one prompt where Hami Tahm was also named as the recommendation, not just cited as a source. | 24 |
The engine breakdown
ChatGPT accounts for 84 of hamitahm.com’s 188 citations, Perplexity 70, Google 19, and Microsoft Copilot 15. Across all 159 recorded citation observations, the average cited position was 5.4 and the median was 4: consistently a source in the answer, rarely the very first one.
Where it doesn’t show up: named recommendations
Out of 39 prompts, hamitahm.com was cited as a source on 10. It was named as the actual recommendation, the AI’s answer explicitly saying to hire this practice, on exactly 1: “Can you recommend an AI visibility consultant in Toronto?”, on Perplexity, on 2026-08-25, at sentiment +50. Every other prompt in the panel, including the generic, high-volume ones like “best AI visibility consultants in Canada” and “AEO implementation guide for transaction-focused landing pages,” produced a citation, a competitor’s name in the answer, or both, but not this practice’s name.
Otterly’s separate Brand Ranking table (the filtered 13-prompt view described above) puts this in a leaderboard against the same named competitors:
| # | Brand | Sentiment | Mentions | Coverage |
|---|---|---|---|---|
| 1 | Longhouse | +61 | 33 | 6.4% |
| 2 | Canopy Media | +48 | 17 | 3.3% |
| 3 | First Page Sage | +70 | 8 | 1.6% |
| 4 | Parabolic Studio | +72 | 7 | 1.4% |
| 5 | Digital 6ix | +57 | 6 | 1.2% |
| 6 | iPullRank | +20 | 6 | 1.2% |
| 7 | Valory | +43 | 3 | 0.58% |
| 8 | Hami Tahm | +50 | 1 | 0.19% |
| 9 | AEO Agency Canada | +50 | 1 | 0.19% |
| 10 | CAUSE Marketing | +100 | 1 | 0.19% |
04: What this isn't claiming
- It does not claim Hami Tahm is Canada's top, or top-5, AI visibility consultant. This measures citation volume and mention count in one 39-prompt panel, not market share, client roster size, or quality of work.
- It does not claim being cited the most is the same as being recommended the most. This post's whole point is that those two numbers moved in opposite directions for the same site in the same window.
- It does not claim this holds outside Canada. Every prompt in this panel is Canada-scoped.
- It does not claim this is permanent. AI answers change week to week; this is a one-week window, not a standing rank.
- It does not claim Otterly's crawler, or its 13-prompt filtered dashboard view, is the ground truth. Every figure above that comes from the full 39-prompt export is labeled as such; the Brand Ranking numbers are labeled as coming from the filtered view instead, on purpose.
Curious where your own brand ranks?
AI Visibility Audit runs the same kind of engine-by-engine competitive comparison shown above, for your category, checked by a person, not just pattern-matched. $1,500 Flat fee, one-time.
Book an AI Visibility Audit →05: Frequently asked questions
Does this mean Hami Tahm is Canada's top AI visibility consultant?
No, and this post doesn't claim that. It's a snapshot of one 39-prompt panel, run over one week, through one tracking vendor (Otterly.AI). Within that specific, narrow scope, hamitahm.com was the single most-cited domain out of 1,046 domains that appeared anywhere in the panel. That is a real, sourced result about citations. It is not a claim about who AI recommends most, which the same data shows is a separate and much weaker number: see the next question.
If the site is cited the most, why isn't Hami Tahm 'recommended' the most?
Because a citation and a named recommendation are different things, and this dataset shows they don't move together. Being cited means an AI engine's answer linked to a page as a source, 188 times here, more than any competing agency's own domain. Being named means the answer's text actually says "hire this person/company," which happened on exactly 1 of 39 prompts. AI engines are already pulling facts and structure from this site; they aren't yet consistently concluding "so this is who you should hire" from what they find. That gap, not a visibility problem, is the real finding here.
Which prompt actually produced a named recommendation?
"Can you recommend an AI visibility consultant in Toronto?", on Perplexity, on 2026-08-25. It's a specific, local, high-intent phrasing, not a generic "best AI visibility consultant" query. Whether that pattern (specific + local beats generic + national) holds up is worth testing further, and this post says so rather than overstating one data point into a rule.
What's the difference between this and the HomeCalc AI search benchmark?
Different subject, same method. The HomeCalc post benchmarks a mortgage-calculator product against Canadian banks and rate sites. This one benchmarks Hami Tahm's own consulting practice against named AI-visibility and AEO agencies in Canada. Both use Otterly.AI, both use the same sum("Times cited") counting method, and neither dataset is ever merged with the other or with the Bing/Google Search Console numbers behind the case studies.
Can I run this same benchmark for my own business?
Yes, that's most of what the AI Visibility Audit does: the same engine-by-engine, prompt-by-prompt competitive comparison, for your category. Running it on my own name first, and publishing the unflattering part, is the point: I'm not asking a client to trust a method I haven't pointed at myself.
For the full strategy framework behind closing a gap like this one, see the AI visibility hub. To see the audit this method is drawn from, see the AI Visibility Audit page.
Hami Tahm is an AI visibility consultant based in Toronto, serving businesses across Canada.
Disclosure: hamitahm.com is this practice's own site, and this article describes a service sold on it. The ranking data itself is unedited: every figure is counted from the raw Otterly.AI export, including the part (8th of 10 by named mentions, not 1st) that is less flattering than a rounder claim would be.