A 48-finding audit, then machine-readable fixes
An audit method that scores a site across search, answer-engine, generative-engine, and conversion readiness at once, proven first on our own business’s real search performance.
Era 6, the AI-agent-native practice · In progress
The situation
Ranking well in traditional search does not guarantee anyone finds you, and it does nothing for the growing share of people who now ask an AI assistant a question instead of typing it into a search box. Our own notary business was ranking in position one on some searches and still getting almost no clicks, a sign the underlying visibility and machine-readability of the site had real gaps beyond ranking alone. That gap is exactly what the audit was built to find and measure, before a single fix was attempted.
What we built
An audit methodology that scores a site across four dimensions at once: traditional search (SEO), answer-engine readiness (AEO, how well a page answers a direct question the way voice assistants and featured snippets expect), generative-engine readiness (GEO, how citable and fact-clear a page is to an AI system generating an answer), and conversion (how clearly the page turns a visitor into a booked appointment). On top of the audit, we built registry-driven structured data (a single machine-readable source of facts about the business that every page pulls from, instead of separately maintained tags per page) and machine-readable surfaces, llms.txt and facts.json, files written specifically so AI crawlers and answer engines can read a business’s facts directly rather than guessing from prose.
Named approval gates
Every fix identified in the audit was logged as a tracked finding before any code changed, and each deploy went through the same health-gate and automatic-rollback pipeline used across the rest of the business, so an SEO fix could not silently break the live site. A single hard rule was applied without exception: only verified metrics with a named source ever reach the public site.
Verified results
- 48audit findings identified (2 critical, 18 high-priority)audit, 2026-07-08
- 41 / 49 / 27 / 45baseline SEO / AEO / GEO / Conversion out of 100V2 scorecard, 2026-07-09
- 4 clicks / 90 daysdespite position-1 rankingsaudit, 2026-07-08
- 14 of 254pages indexed by Bing at audit timeaudit, 2026-07-08
The initial audit identified 48 findings, 2 critical and 18 high-priority, including a Google Business Profile listing that was effectively invisible to searchers and a Bing index covering only 14 of 254 pages. A baseline scorecard measured the site at 41/100 SEO, 49/100 AEO, 27/100 GEO, and 45/100 conversion, giving a fixed starting point every later change is measured against. The site was getting 4 clicks in 90 days on searches where it already ranked first, the specific gap the registry and machine-surface work targets.
What this means for your business
A page can rank and still be invisible in the ways that matter now: to the AI assistants your customers increasingly ask instead of Google. We audit against all four dimensions at once, on a baseline you can see the numbers for, and we proved the method on our own business’s real search performance before offering it as a service.