What actually counts as ranking signal in Canada (2026)
After auditing 80+ Canadian local pack engagements between 2023 and 2026, we have a stable picture of what moves the needle. The split is roughly 60% Google Business Profile signals and 40% on-site signals, with NAP/citation hygiene as a baseline rather than a lever.
| Factor | Approx. weight | Lever or baseline? |
|---|---|---|
| GBP primary category match | High | Lever |
| Review velocity (trailing 90 days) | High | Lever |
| Owner-answered Q&A in last 90 days | Medium-high | Lever |
| Named-neighbourhood service pages | Medium-high | Lever |
| Sector-specific schema (LegalService, Dentist…) | Medium | Lever |
| NAP consistency across citations | Low | Baseline |
| Total citation count | Very low | Baseline (don't chase volume) |
| Domain authority of website | Low | Indirect |
Until ~2020 buying citations from the standard Canadian directories (YellowPages, Yelp, 411.ca, Goldbook) moved Map Pack rankings. In 2026 it does not. Maintain consistency in the top ~12 directories, then redirect that budget to review acquisition and named-neighbourhood content.
GBP categories: the single most under-used lever
Most Canadian local pack losses we audit trace back to a wrong primary category. Google ranks the primary category about 5× more heavily than secondaries. Pick the most specific category that exactly matches your highest-revenue service, even if it covers only 40% of your offering.
How to choose a primary category
- 1List your top 5 services by revenue in the last 90 days.
- 2For each, search Google for that service + your city. Look at the businesses appearing in the Map Pack — their primary category is often visible in the knowledge panel.
- 3Pick the category that maps to your top-1 revenue service. Use up to 9 secondaries for the rest.
- 4Avoid generic umbrella categories ('Service Establishment,' 'Business') — they suppress ranking for the specific intent.
Changing primary category temporarily depresses rankings for 7–21 days while Google re-evaluates. Plan a re-categorisation outside peak season for your business — January for HVAC, late November for tax services, March for landscapers.
Review velocity targets by vertical
Review velocity (new reviews per month over the trailing 90 days) is the strongest predictor of Map Pack movement we measure. The targets below are floor numbers — competitors hitting them will outrank you on equal review counts.
| Vertical | Reviews / month (floor) | Reviews / month (winning) |
|---|---|---|
| HVAC, plumbing, electricians | 10 | 20+ |
| Law firms, accountants | 4 | 8+ |
| Dental, medical clinics | 8 | 15+ |
| Restaurants (independent) | 12 | 25+ |
| Auto repair, collision | 10 | 20+ |
| Real estate brokerages | 6 | 12+ |
| Med-spa, beauty | 10 | 20+ |
How to actually hit these: build review collection into the moment of value delivery (technician's tablet at job close, hostess at table-clear, dental hygienist at post-cleaning). Email campaigns to the database have a ~2% response rate; in-person ask at moment-of-value is 25–40%. The ROI difference is enormous.
Named-neighbourhood pages, not postal-code pages
Build pages for the 5–15 named neighbourhoods you actually serve, with content distinct to each. Toronto SEO's audit set shows named-neighbourhood pages ('Liberty Village dentist,' 'Plateau Mont-Royal accountant,' 'Kitsilano renovation contractor') outperforming postal-code pages by 4–6× on local pack inclusion.
Each neighbourhood page should answer questions specific to that area: housing stock, demographic, common pain point, local landmark, parking situation. A 'we serve Leaside' page that mentions post-war bungalows, Bayview, the LRT, and 1940s knob-and-tube wiring is a different page than a generic city page with the neighbourhood name swapped in.
Generating 200 pages for every Toronto FSA (M4K, M5R, M6P…) with templated copy is a textbook 'thin content' pattern. Google flags it within 6–12 weeks and the entire batch gets demoted, often dragging the parent domain with it. Do not do this.
The bilingual Quebec edge most agencies miss
Quebec searchers default to French in Google. A French-language GBP profile, a /fr landing page, and bidirectional hreflang on the English equivalent unlocks a market most Canadian SMBs leave on the table. In our Quebec engagement set, the French-language Map Pack has substantially less competition than the equivalent English-language pack for the same query.
Operationally: create a separate GBP profile for your Quebec location with French primary category and French business description; add a /fr landing page with native-French copy (not machine-translated); and add hreflang tags on both pages. A native French-speaking reviewer is essential — Quebec searchers detect machine-translation immediately and bounce.
Quebec's Bill 96 (in force 2023, with deeper requirements landing in 2025–2026) requires French-language signage and customer-facing material for businesses operating in Quebec. A French-language website is increasingly a compliance question, not just a marketing question.
How AI Overviews pick local businesses to cite
Google's AI Overview for a local query (e.g., 'best dentist in Mississauga') typically cites 3–6 businesses. The selection mechanics are different from the Map Pack — AI Overviews favour businesses whose website provides extractable, fact-dense content that answers the implicit sub-questions of the query.
- 1Have an /about page with named team members, credentials, and provincial regulator number.
- 2Use sector-specific schema (LegalService, Dentist, HVACBusiness) instead of generic LocalBusiness — the named type signals what kind of business you are at parse time.
- 3Have a clear pricing or 'how much does X cost in [city]' page with current ranges. AI Overviews answering pricing questions cite businesses that publish ranges far more than businesses who hide them.
- 4Include hours, payment methods, accessibility, and languages in your visible content (not just GBP). Microsoft Copilot and Perplexity scrape these from on-page content as well as GBP.