What we do
A defensible discovery baseline — not a “visibility score”.
For doctor owner-operators and med-spa buyers: we measure what a future patient (or referrer) asks when they’re choosing a provider, show what the model actually said (with provenance), and map concrete fixes you can ship this month—followed by a comparable verify rerun.
Discovery prompts (primary)
Category + geo questions first ("best wound care in Sarasota", "how to choose a home health agency near Lakewood Ranch"). Brand-aware checks come second.
Truth boundaries
This is low‑n by default. Prompt choice matters. We show provenance (seed/coverage pack, prompt_set_hash, provider/model, methodology_version) and only treat seed reruns as comparable when those match.
Repeatable loop
Run → implement fixes → verify rerun. Trend is only meaningful for seed baselines when pack/hash/provider/model/methodology stay aligned.
How this complements Google/Meta
Ads and SEO still matter. We focus on the public signals AI tools retrieve and quote—often the same signals that improve conversion and local search. The output is a fix list you can hand to your web team, not a slide deck.
Fast vs slow changes
Some fixes move quickly (page clarity, service coverage, citations). Others take time (authority, competitive landscape, product changes). We label expectations and verify only when runs are comparable.
What we don’t do
No guarantees. No universal claims. We don’t pretend low‑n is a statistically rigorous survey. We show provenance so stakeholders can interpret results correctly.
Deliverables
Baseline evidence pack, prioritized fixes, and a verify rerun plan tied to a stable prompt pack + model. For buyers, this doubles as due diligence signal on “discoverability without paid clicks.”