What every number in the audit means, what moves it, and how to answer the owner's questions. Nothing here is theoretical: each section names the exact measurement behind the number.
One weighted composite of five measured components:
| Component | Weight | What we actually measure | |---|---|---| | AI answers | 30% | Of ~24 valid answers from ChatGPT, Perplexity, Gemini and Claude to 6 homeowner questions ("best roofer in {city}"…), in how many is THIS business named? | | Local map pack | 25% | For those same 6 searches on Google (mobile, their city): does the business appear in the 3-slot map pack? | | Organic results | 15% | Same 6 searches: does their website appear in the top 20 blue links? | | Google Business Profile | 15% | Completeness /100: categories, hours, site link, description, services, photos, Q&A, review responses. | | Site performance | 15% | Google PageSpeed mobile score of their current website (no site = 0). |
Worked example — Amity Roofing, 12/100: AI mentions 0/24 → 0 of 30 pts. Map pack 0/6 → 0 of 25. Organic 0/6 → 0 of 15. GBP 80/100 → 12 of 15. Site speed 60/100 → 9 of 15 (weighted ≈ 12 total). Translation for the owner: "Your Google profile and your reviews are good. Everything that brings strangers to you is broken."
- 0–15: invisible. The business exists only for people who already know its name. Strongest pitch: pure invisibility + receipts. - 16–30: barely visible. Usually one component works (often GBP or a few map-pack hits). Pitch: "you're leaking in 4 of 5 channels." - 31–50: partially visible (e.g. Toitures de l'Estrie, 33 — IN the map pack but zero AI presence). Pitch flips: "Google found you. AI hasn't — and that's where search is going." - 51+: rare in our lead pool; if seen, qualify harder before pitching.
- AI answers: AI assistants cite businesses with crawlable, structured websites and lots of recent review text. Our drop provides the structure; their review growth provides the evidence. Moves in weeks-to-months, not days. NEVER promise a timeline. - Map pack: proximity (fixed), relevance (GBP categories/services — one-time fix), prominence (reviews volume/velocity/recency + citations). Review VOLUME and RECENCY are the levers we influence most. - Organic: needs a real, fast, indexed website with per-service pages. The drop is this, wholesale. - GBP: mostly a ONE-TIME completeness fix (done at claim) + light upkeep. - Site performance: fixed permanently by the drop's architecture.
"Where do these numbers come from?" Every one is measured, dated, and listed in the methodology footer. The AI quotes are verbatim transcripts. If he wants, he can type the same question into ChatGPT on the call.
"My phone rings fine." "From referrals and repeat customers — the people who already know you. This score measures the strangers. 280 measured searches a month in your area are meeting your competitors first."
"Can you get me to #1?" "No one can promise that, and whoever does is lying. We guarantee the inputs — the site, the structure, the speed, the profile — and we report the outcomes monthly, with numbers like these."
"Is AI search even real?" "You just saw Claude recommend Blue Hammer by name. Every iPhone ships with AI answers now. This is where 'Google it' went."
"Why is my score low if my rating is 5.0?" "Because rating measures how good you are; this measures how findable you are. That gap is exactly the problem — and it's the fixable half."
- Quote ONLY numbers that appear in the audit document. - needsReverify banner on the pack = do NOT use "your website is broken" hooks (verdict unverified). - Exposure dollars are a RANGE with printed assumptions — say "estimated", never promise it as income.