AI Visibility Index · Greater Victoria · October 2026 pilot
Which Victoria accounting firms does AI name?
We asked ChatGPT, Gemini, Google AI Mode and Claude the same ten small-business accounting questions and recorded which local firms each one named. The biggest finding: every AI tool told a different story. Only 6 of 15 local firms were named on all four.
Research preview. Measures name presence, not service quality.
Key findings
One AI tool is not the whole market.
6 of 15
local firms were named on all four AI tools tested.
40
AI answers captured: ten questions on each of four tools.
17
answers named Baker Tilly Victoria and Padgett Victoria, tied for the highest presence in this local sample.
1 of 4
tools named Sterling Stanford and RMC CPAs in this captured sample; the other three sources did not name them.
The results
Firms named across the pilot sample.
Each AI tool received the same ten questions. One answer can name several firms, so shares don’t add up to 100%. Coverage shows how many of the four tools named a firm at least once.
| Sample rank | Firm | Name presence | Coverage | ChatGPT* | Gemini* | AI Mode* | Claude |
|---|---|---|---|---|---|---|---|
| 1 | Baker Tilly Victoria | 17/40 | 4/4 | 1 | 6 | 5 | 5 |
| 1 | Padgett Victoria | 17/40 | 4/4 | 5 | 6 | 5 | 1 |
| 3 | Prosper CPA | 15/40 | 4/4 | 5 | 3 | 5 | 2 |
| 4 | Avalon Accounting | 13/40 | 4/4 | 2 | 1 | 3 | 7 |
| 4 | Ralph & Levinson | 13/40 | 4/4 | 2 | 6 | 4 | 1 |
| 6 | Modern Axis CPA | 12/40 | 3/4 | 0 | 3 | 2 | 7 |
| 7 | Burkett & Co. | 11/40 | 3/4 | 0 | 4 | 6 | 1 |
| 8 | Dusanj & Wirk | 8/40 | 4/4 | 1 | 3 | 2 | 2 |
| 9 | Hutcheson & Co. | 7/40 | 3/4 | 0 | 3 | 1 | 3 |
| 9 | MNP Victoria | 7/40 | 3/4 | 2 | 1 | 0 | 4 |
| 9 | Dickson CPAs | 7/40 | 2/4 | 0 | 4 | 3 | 0 |
| 12 | RMC CPAs | 5/40 | 1/4 | 0 | 0 | 0 | 5 |
| 13 | Schibli Stedman King | 4/40 | 3/4 | 0 | 1 | 1 | 2 |
| 13 | Hampton & Company | 4/40 | 2/4 | 1 | 0 | 0 | 3 |
| 13 | Sterling Stanford | 4/40 | 1/4 | 4 | 0 | 0 | 0 |
Coverage = how many of the four AI tools named the firm at least once. * Search Atlas Prompt Simulator observations. Claude observations were collected manually from the Claude Pro web app with chat memory disabled. Ten prompts per source, 40 source-question observations in total. This table covers 15 selected Victoria-area firms, not every company discovered. Sample ranks compare these 15 selected local firms by the number of answers naming each firm. Equal totals share competition ranks (1, 1, 3); coverage and name determine display order within a tie. Other captured firms are outside this selected table. These are observed presence counts, not service-quality scores.
Coverage matters
Total mentions can hide a blind spot.
Dusanj & Wirk
Named in 8 of 40 captured answers, with at least one appearance on each of the four tools.
Ralph & Levinson
Named in 13 of 40 captured answers across all four tools after matching the former name C. Ralph & Associates to the merged firm. The Claude appearance used that former name.
Sterling Stanford
Named in four captured ChatGPT answers and absent from the captured answers on the other three tools.
Serving vs based here
Two providers market to Victoria from elsewhere.
Their mentions are part of the captured evidence, but they are shown separately from the local table above.
| Provider | Based in | Name presence | ChatGPT* | Gemini* | AI Mode* | Claude |
|---|---|---|---|---|---|---|
| WTC Chartered Professional Accountant | Toronto, ON | 6/40 | 0 | 3 | 3 | 0 |
| Hayson Accounting | Whistler, BC | 4/40 | 1 | 2 | 1 | 0 |
Method
What readers need to know.
How was the pilot presence rate calculated?
A firm’s presence rate is the number of captured answers that named it, divided by 40. Example: Baker Tilly Victoria = 17 ÷ 40 = 42.5%. It can be compared within this pilot, but it is not a production recommendation score. No citation or authority score is assigned.
What questions were asked?
Ten business-intent questions about incorporation, bookkeeping and year-end, payroll and GST, professional services, QuickBooks, business structure, advisory, retail, trades and construction, and cloud accounting.
How were the answers collected?
Search Atlas Prompt Simulator returned 30 completed answers from ChatGPT, Gemini and Google AI Mode. A separate Claude Pro rerun with memory off produced ten more observations on 18 October 2026.
How big is the sample?
One response per question per tool. AI tools rarely give the same answer twice, so some differences may be normal run-to-run variation. Future editions will use repeat samples, and larger samples may change the order.
How were firms identified?
Names and clear aliases were consolidated, and each firm was counted at most once per answer. Official websites were checked for identity and a Victoria-area presence. This is an identity and presence check, not verification of every claim AI made.
What was not measured?
Citation frequency by firm, sentiment, positive recommendation rate, ranked position, citation-domain strength, client experience and CPA quality.
Editorial check, 21 October 2026
The 15 local firms were recounted against the same 30 captured Search Atlas answers and ten controlled Claude answers; no new scan was run. Padgett includes CountBeans / Count Beans (17/40). Modern Axis includes ModernAxis (12/40). Ralph & Levinson includes C. Ralph & Associates, which merged into the firm in 2025 (13/40). Overlapping names within one answer count once. This changes all-source coverage to 6 of 15 local firms. MNP remains 7/40; Hayson and WTC remain in the remote-provider section. Presence counts include incidental mentions and places panels.
What’s next
The first step toward a repeatable Index.
Victoria is our pilot market. We’ll run the Index regularly across more local industries, starting on Vancouver Island and expanding to Vancouver. Each edition adds deeper measures, like recommendation prominence, share of voice and citation evidence, so results can be compared over time.
Is your business named on every AI tool?
The free AI Business Diagnostic shows how visible your business is to AI, in about five minutes.