Questions we get before the first call.

Getting your business named. Ideally first. When someone asks ChatGPT, Gemini, Perplexity or Google's AI Overviews a question you should be the answer to. It's measured in share of answers, not rankings or traffic.

SEO earns you a position on a results page. Answer engines don't show a page. They name a handful of businesses. Most of what decides that happens off your website, in reviews, roundups, forums, video transcripts and local coverage. Schema and FAQ blocks are the cheap half of the job.

You see the full prompt set on day one and can run any prompt yourself, any evening, in the same chat window your customers use. Every monthly report includes the raw answers so you can read what the model said about you word for word.

Five ways, all built in. Every prompt has one-tap buttons that open it in ChatGPT, Claude, Perplexity and Google AI Mode so you can compare the real app's answer to ours. Every stored answer lists the sources the engine read, with the model version and the time stamped on it. Each prompt is asked three times a week per engine, so you see a rate ("named 2 of 3"), not one lucky sample. You can re-run any prompt live from the dashboard while you watch. And you or your team can log what the real app showed; the dashboard keeps a running agreement score between the apps and our data. Full methodology.

You get a live dashboard per location. It shows share of answer. How often you're named, named first, and cited across the prompts that matter. Updated from weekly prompt runs, next to AI-referred visits and form fills from your own analytics. Every raw answer is kept, so you can read what the model said.

ChatGPT, Gemini, Perplexity and Google AI Overviews. Prompts are re-run weekly and reported monthly, per location.

No. The audit and the build are fixed-scope, one-time fees. The monthly authority retainer is month to month and can be cancelled with 30 days' notice. You own everything we build.

No. We take one client per category per metro. If we're working with a dispensary in San Diego, we won't take another one.

The audit ships in 7 days and the build in 10 working days. Off-site mention work starts in month one, and the before/after comparisons we show are framed around a 90-day window. Individual results vary by market and category.

There's no public search-volume data for AI prompts, so we don't invent any. The set is our read on which questions actually produce a recommendation: discovery, first-timers, comparisons, price and deals, and logistics like hours or delivery, each tied to a location. You see the whole set on day one and every prompt opens in the real apps.

Across 75,000 brands, YouTube mentions had a 0.737 correlation with AI visibility, the strongest single signal measured; backlinks scored 0.218. Subscribers made no difference, description length was the strongest metadata signal, only a third of cited videos had chapters, and 94% of citations went to long-form video. Foundational video datasets were built on YouTube transcripts, so models already treat it as a knowledge source. These are correlations, so we treat video as a strong bet with cheap downside, not a guarantee. For a multi-location retailer it's unusually winnable: a store walkthrough filmed on a phone, transcribed and chaptered, needs no audience.

Because we had the problem before we had the service. We watched customers arrive already holding a recommendation and realised it wasn't ours. Every tactic is tested across our own locations before a client is billed for it, and our own live numbers sit on this page, weak locations included. We also know what a review programme does to staff morale and why a deliverable that needs six hours of the owner's time never gets done, so the plan is written around the operation, not a reporting template.

You run it yourself, on this page, in about twenty seconds: three real questions for your business and city, asked to ChatGPT and Gemini with web search on. You see the actual answers with your name highlighted. Leave an email and the same results land in your inbox within a minute, with a link you can send to a partner. Ask and we will run the deeper ten-question, four-engine version by hand. No call attached either way.