AEO · Original finding

Roughly 41% of Canadian SMB sites we audited in Q1 2026 still block at least one major LLM crawler (GPTBot, ClaudeBot, PerplexityBot) by accident in robots.txt — typically inherited from a CDN default.

Across 80 Canadian SMB site audits in Q1 2026, 33 (41%) had at least one LLM crawler blocked in robots.txt. The most commonly blocked were GPTBot (28 sites) and Amazonbot (24 sites). In every case the operator confirmed the block was unintentional and inherited from a Cloudflare or default-provider configuration. Sites that opened the block saw measurable LLM citation appearances within 4–8 weeks.

Methodology

robots.txt audit across 80 Canadian SMB sites, Q1 2026; follow-up citation probe at 8 weeks for the 18 sites that opened the block.

Limits of this finding

Allowing crawlers does not guarantee citation; it only removes a blocker. Citation lift was strongest on sites with original research or pillar guides already published.

Run experiments like this on your site.

Senior-led measurement plans, hand-counted, no laundered numbers.

Book a strategy call