Measurements, not opinions, on how AI answer engines decide which brands to name. Everything here comes out of the same pipeline we run for customers: fixed prompt sets, nine engines, five runs a day, full answer text kept.
Where a post makes a claim about how an engine behaves, the source is linked. Where a number comes from our own data, the methodology is published alongside it.
An entity is a resolved, machine-readable identity distinct from your website. Without one, engines read every mention of your name as an unlinked string.
Semrush's headline metric is one blended AI Visibility Score. Six dedicated trackers compared on price, prompt count and per-engine reporting.
Three different outcomes with three different fixes. A mention can come from training memory, a citation needs a live fetch, and only one of them sells.
A drop across every prompt on one engine is a model or index change. A drop on a few prompts across every engine is a source of yours…
Six AI visibility tools priced under $200, compared on the unit each one meters. Prices read from every vendor pricing page on 17 August 2026.
AI share of voice is the share of answers in a fixed prompt set naming your brand, split across named competitors. The method, and what moves it wrongly.
Three numbers get misread in every AI visibility report: a mention taken as a recommendation, one run taken as a rate, a branded prompt taken as reach.
Peec includes three models per tier and bills extras per model, with one project at entry. Six alternatives compared on total cost at real coverage.
Bing Webmaster Tools reports Copilot citations, the pages cited, and the queries used to retrieve them. It is the only free first-party AI diagnostic.
Otterly includes four core engines and bills the rest per engine, so compare the total at the coverage you need. Six alternatives, every price dated.
Gemini grounds in Google’s index, so Search eligibility is the entry ticket. That direct relationship is unusual, and a bad proxy for the other engines.
A minimum viable prompt set covers four intents. Most audits test only the first and overstate the result. Here are the twelve prompts, ready to copy.
