What gets measured in each type of business
AI visibility is not measured the same way for a law firm as for an online shop. Here is what changes by sector.
Professional services
Law firms, consultancies, accountants, agencies.
What decides a hire here is not the product question, it is the problem question: nobody searches for "employment law firm", they search for "I have been fired and I think it was unfair".
This is where measuring by intent instead of by service shows up most clearly. The sources AI tends to cite are professional directories, trade media and forums — almost never the firm's own site, which is exactly the problem to solve.
Health and clinics
Local intent combined with symptom intent dominates, and trust signals weigh more than in any other sector.
We check that the listing has an address, opening hours and ratings a machine can read, because those are the signals a model uses to decide who to name when the question includes a city. Measurement is split by specialty, not by clinic: showing up for "dentist" but not for "invisible braces" are two different problems.
E-commerce
Your competition here is not other shops: it is comparison sites and recommendation media, which take most of the citations.
The useful questions are category and comparison questions ("best X for Y", "cheap X that lasts"), not brand ones. We check that product pages carry price, availability and ratings in a format a machine can read.
Software and SaaS
The battleground is comparisons and alternatives: "alternatives to X", "X vs Y", "best tool for Z".
This is where the Library Effect shows up most plainly: models know an enormous amount of software and recommend very little of it. The deciding sources are usually review sites, technical communities and third-party comparisons.
Industry and manufacturing
Technical questions, long cycles and very few sources: the sector with the least competition for the citation and therefore the cheapest to win.
Measurement goes by application and specification, not by product name. With little editorial coverage around, one well-built technical page can become the source the whole market cites.
Education
Outcome intent rules: not "course in X", but "how to change careers" or "what to study to work in Y".
In this market AI answers tend to list programmes and schools together, so Share of Mentions says more than Share of Answer: what matters is the slice of the total you take, not just whether you appear.
Hospitality and travel
Everything is local and occasion-driven: "where to take clients for dinner", "quiet hotel near", "place to celebrate something".
Citations almost always go to guides, local media and aggregators. We check that the listing has a menu, hours, address and readable ratings, and we measure by occasion rather than by venue type.
And if your sector is not on this list That is fine: the list is only the set of types the engine recognises in order to pick a starting point. Questions are built on your real verticals, and you can correct them. If the type cannot be determined from the website, the report says so instead of quietly applying a generic template.
See what comes up in your sector
The free audit uses the questions from your market and tells you who shows up instead of you.
Run my free audit →Home · Plans & pricing · Methodology · Metrics · How to choose · Changelog