Which AI assistant Canadians use most, and the engine behaviour that should change your spend instead

Usage share is the wrong input for this decision, and it is what every vendor leads with. The right inputs are two behaviours we measured directly on Canadian questions. How often each engine reads the live web, because that decides how fast your publishing takes effect and when a re-read is fair. And how often each engine commits to naming a supplier, because that decides whether the slot in your category is occupied or empty. On 22 October 2026, across Canadian buying questions, ChatGPT ran a live search on all 32 of its readings and Gemini on 20 of its 32, while Gemini named a rival in 17 readings against ChatGPT's 13. Those two rows should change a budget. A usage share percentage should not.

What we measured, per engine, on Canadian questions

Reading, Canada, 22 October 2026	ChatGPT	Gemini
Blind readings, brand batch excluded	32	32
Ran a live web search	32	20
Named a supplier that was not us	9	17
Named no supplier at all	23	15
Named us	0	0

The bank was 198 Canadian buying questions in 33 batches, put to both surfaces through a data provider with Canada named inside the text of every question, because the provider accepts gl, hl and location parameters and ignores all three. Batch 33 names us in the question and is a ceiling measurement, excluded throughout. The two engines produced 2,992 citation URLs across 974 distinct hosts between them, and they are reported separately rather than averaged, because the live search row is exactly what an average destroys.

The reversal, and why it matters to a Canadian buyer

We ran the same bank for the United States on the same day. There ChatGPT named a rival in 13 of its 32 readings and Gemini in 11 of its 30, and ChatGPT ran a live search on all 32 while Gemini ran one on 15 of 30. So the engine that named more rivals flipped between the two markets: Gemini in Canada, ChatGPT in the United States. Two separate measurements with their own denominators, reported side by side and never merged.

The consequence for a Canadian company is direct. A global benchmark, a global usage share or a global visibility score would all have missed this, and a programme that optimises for whichever engine leads a worldwide usage table would be optimising for the wrong one in Canada. Measure your own market.

The two behaviours, and what each one should change

• Live search frequency sets your re-read schedule. On Canadian questions one engine read the live web every time and the other on roughly two thirds of readings, so a page published this week reaches them on different cycles. A fair re-read date therefore differs by engine, and a blended figure at four weeks measures two different lags as one.
• Naming frequency sets where you start. ChatGPT declined to name anybody in 23 of its 32 Canadian readings and Gemini in 15 of its 32, which is 33 of 64 in total and the largest single outcome. Those readings are the cheapest openings, because an unoccupied slot needs a published answer rather than the displacement of an incumbent.

Why usage share is a poor purchase input

Three reasons, each sufficient. The published share figures disagree because they define the category differently, with the biggest disagreement being whether an AI answer inside a search results page counts. Share says how many people might ask and not what they are told, and the second is the part you can change. And share is a national average while buying behaviour is category specific, so a Canadian figure can be accurate and irrelevant to your market at once.

What share is good for is a rough ordering when you can only afford two engines. Five surfaces are sold in this category, ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews, and reading two honestly while naming which two is a real measurement. Reading one and calling it AI visibility is not.

What the engines cited in Canada

Across the 64 blind Canadian readings the most cited hosts were developers.google.com in 17 readings, support.google.com in 14, torontoseo.com in 13, clutch.co in 12, canadacreate.com in 11, canada.ca in 10, wisemedia.io in 9, semrush.com in 9, competition-bureau.canada.ca in 9, ised-isde.canada.ca in 9, blog.google in 8, www150.statcan.gc.ca in 8, ottawaseo.com in 7, searchengineland.com in 6 and agencies.semrush.com in 6. One entry, ccn.web.tr, is the data provider's own fetch infrastructure rather than a publisher: it is excluded from every source analysis and never merged into another host. Where a supplier was named there were 48 software namings, led by Semrush at 16, HubSpot at 12 and Ahrefs at 9, and 21 namings of six Canadian agencies.

How to decide which engines to pay to read

• Read at least two and name them in every report. One engine is an anecdote with a dashboard.
• Prefer the pair that behaves most differently, because that is what reveals whether a result is about your business or about one engine's refresh.
• Do the arithmetic before you look at plans. Questions times engines times frequency times markets times language sets. The engine count is a multiplier, so the fifth engine costs more than the second.
• Ask whether the vendor reports the engines separately. A blended score would have hidden both Canadian findings above.
• Ask whether you get the raw answer text, because a score with nothing under it cannot be audited.

If Gemini is the engine you are least sure about, the vendors that treat it as an extra rather than as standard are identified in two vendors sell Gemini as an extra. If the surface in question is Google's AI Mode rather than Gemini, the state of tooling for it is in a surface we do not read and the two vendors that claim it.

Questions about which assistant to care about in Canada

Which one do Canadians use most?

The published share figures disagree on definition and we did not measure usage. What we measured is behaviour on Canadian buying questions, where ChatGPT read the live web on all 32 of its readings and Gemini named rivals nearly twice as often.

Should I optimise for the biggest one?

No, because the behaviours differ and one engine's result does not predict the other's. In Canada the more rival naming engine was the opposite of the market next door.

Does the leader change by category?

Very likely, which is a further reason a national average is a poor input. Your own reading on your own questions is the category specific answer.

Why does live search frequency matter so much?

Because it sets how quickly a published page can appear in an answer, and therefore when a re-read is fair. Two engines with different frequencies need different re-read dates.

Is it worth paying for all five surfaces?

If your Canadian buyers use them, yes, and the cost rises with the engine count because engines are a multiplier. Start with two named engines and add deliberately.

Reading both rather than betting on one

The decision this page argues against, picking a leader and treating its answer as the answer, is what our own Canadian table disproves: the two engines differed on live search by half again and on naming by nearly double, in opposite directions to the market next door. AI Knows Us reads ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews with Canada named inside the question, reports each engine separately with its own denominator, and keeps the raw answer text. Outside the United States and the United Kingdom the cheapest tier that reads all five engines is Growth at 199 dollars a month with 100 buyer questions watched, read on aiknowsus.com/pricing and aiknowsus.com/llms.txt on 22 October 2026. Outbound, at 869 dollars in that column, reads four of the five rather than all five, which is the kind of detail worth checking on any plan.