Home/AI Visibility

Flagship service

AI Visibility

Get named inside the answers ChatGPT, Perplexity, Google AI Overviews and Claude give your buyers. Measured with a 72-check audit, fixed in priority order, tracked every month.

Get my free first look See the 72 checks

A consultant presenting a printed scorecard to a business owner in a Toronto office

The problem

Your buyers stopped scrolling. They ask.

A growing share of buyers open an answer engine, ask a full question and take the one reply. That reply names one, two or three businesses. If it does not name you, the click, the call and the sale never happen. Search rankings do not fix this. Being quotable does.

An answer engine works like a careful buyer in a hurry. It finds the pages that answer the question, checks that the facts agree across sources, prefers the ones that are recent and specific, and writes a short answer with the names it trusts. Every one of those steps has a failure point that has nothing to do with how good your business is: a crawler blocked at the CDN, a name spelled three ways across the web, a service page that never states the answer, a review profile nobody replies to.

AI visibility is the work of removing those failure points, in the order that moves the number most. It starts with a measurement, because guessing at fixes is how agencies burn a year of retainer.

What you get

One audit. Three reports. One ranked list.

Report

The scorecard

Overall score and grade, ten pillar scores, all 72 checks with one observed fact each. Pass, needs work or fix first, never a guess.

Report

The prompt-test appendix

Five to eight buyer questions run live on four engines, several times each. Every business named, in order, with what was said. You against three competitors as a share-of-voice matrix.

Plan

The ranked fix list

Every failing check ranked by weight times lost score. The top of the list is always the item worth the most. The first month is the cheap, high-impact work.

Baseline

The re-audit file

Stored prompts and scores so the next audit measures the same things. Deltas per check, per pillar and overall, so movement is real, not judgement drift.

The 72 checks

Ten pillars, weighted.

Each pillar carries a weight out of 100. The score reflects what decides whether an engine names you, not a flat checklist. The full explanation of every pillar is on the home page; here is the shape.

AI answer-engine presence and mentions (16). Whether the engines cite you, how often, how prominently, and how that compares with competitors.
Content answerability and quotability (14). Answer-first structure, question headings, extractable passages, real lists and tables, FAQ blocks, depth, freshness.
AI crawlability and access (12). Bot policy, the live fetch test as GPTBot, ClaudeBot and PerplexityBot, rendering without JavaScript, indexing, sitemap.
Technical SEO and performance (10). Core Web Vitals, load time, mobile, HTTPS, canonicals, redirects.
Structured data (9), entities (9), E-E-A-T (9), on-page (9). The evidence layer: schema, one clear entity, trust signals, the fundamentals.
Authority and off-page (7), measurement (5). Who vouches for you, and whether you can see the number move.

One check, in full"Brand cited across AI answer engines" carries 4 points, the heaviest single check. It scores 100 when you are cited in 80% or more of priority prompts across multiple engines, 75 at roughly half on two engines, 50 at a quarter or on one engine, 25 when cited rarely, 0 when never. Every check has anchors like these, so two people scoring the same site land on the same number.

How it runs

Audit, fix, publish, monitor.

Audit

We agree on the competitors and the buyer questions, then run the full rubric: crawl, schema, SERPs, backlinks, Lighthouse, the bot fetch test and the live prompt matrix. Five business days.

You get: the scorecard, the appendix and the ranked fix list.

Fix

We work the list from the top. Crawler access and entity clean-up first, then answer pages, schema, llms.txt, FAQ blocks and the mentions the engines trust. First fixes live within 30 days.

You get: rewritten pages, schema and crawler access, done for you.

Publish

Changes go live on your site, whether we built it or you already have one. If your platform makes something impossible, we say so and propose the workaround.

You get: a live site the engines can read end to end.

Monitor

Every month we re-run the prompt tests, re-score the pillars that changed and send one plain-English page. Three to six months is where the gains compound.

You get: a monthly share-of-answer report, and a re-audit each quarter.

Timeline and pricing

Start free. Pay for the depth you use.

Free

First look

Two business days. The live half of the rubric plus the three fixes we would do first.

Fixed

Full audit

Five business days. Quoted after the first look, scoped to your site's size and market.

Monthly

Growth retainer

Month to month. Scoped to your goals after the audit. Cancel any time.

Quarterly

Re-audit

Included in the retainer. The same prompts, the same anchors, the delta.

We do not publish shelf prices because a single-location clinic and a national manufacturer are not the same job. What we do publish is the shape: a free first look, a fixed-price audit, a monthly retainer you can leave, and you own everything we make. Ways to work together has the detail.

Questions

Plain answers.

What exactly is in the 72-check audit?

Ten weighted pillars: AI answer-engine presence and mentions, content answerability, AI crawlability and access, technical performance, structured data, entities and brand knowledge, E-E-A-T and trust, on-page SEO, authority and off-page, and measurement. Every check scores 0 to 100 against fixed anchor bands and comes with one observed fact and one fix. You get the scorecard, the ranked fix list, the prompt-test appendix and the competitor share-of-voice matrix as a report you can keep.

How is the free first look different from the audit?

The first look is the live half of the rubric, run in about 20 seconds from the outside: crawler access, structured data, answerability, entities, on-page, technical and agent readiness. The full audit adds the four pillars that need real engine tests and your data: whether the engines actually name you against your competitors, your authority off-site, E-E-A-T across the whole site, and Search Console and GA4.

How long does the audit take?

Five business days from the day you confirm the competitors and the questions. The free first look is two business days.

What does the retainer include?

Month to month after the audit. Each month we work the ranked fix list: rewrite and structure answer pages, ship schema and llms.txt, fix crawler access, build the FAQ blocks, earn the mentions the engines trust, and re-run the prompt tests. You get one plain-English page a month: what we found, what we fixed, what it moved.

Do you guarantee we will be cited?

No, and nobody honest can. The engines change and we do not control them. What we guarantee is a measured baseline, a ranked list, real work against it, and a re-measured number every month so you can see it move or stop paying.

Does this replace SEO?

It includes the parts of SEO the engines still lean on (crawlability, on-page, technical, authority) and adds the parts SEO never measured: whether the answer engines name you, whether your pages are quotable, whether your entity is clear. Most clients keep their rankings and gain the mentions.

From the blog

What is AI visibility? A plain guide for business owners How ChatGPT, Perplexity and Google AI Overviews choose who to cite

Become the answer.

Start with the free first look. Your scorecard lands in your inbox within two business days.

Get my free first look

Or check your site in 20 seconds