1
Run prompts
A versioned panel of buyer-style questions, four AI assistants, and retained repeat observations.
Every prompt is a real question a buyer might ask before hiring — "best NYC divorce lawyer for a spouse with a business", "how do custody rulings across state lines work", "which cosmetic surgeon in NYC does rhinoplasty for men". Atlas sends the frozen prompt panel to ChatGPT, Claude, Gemini, and Perplexity and preserves every completed response. Exact prompt, model, and repeat denominators are published with each snapshot.
You see
A complete transcript. Every response, every name mentioned, every URL cited. Downloadable.
2
Measure visibility
For every prompt: how often were you named? Divide by attempts. That is your citation rate.
Aggregate across the prompt set for your vertical, and you have your share of AI recommendations — expressed with a 95% Wilson confidence interval so you can tell which changes between snapshots are real and which are sampling noise. A rank of #12 with a wide CI is not the same signal as #12 with a tight one; we show you both.
You see
Your rank in the vertical, response mention rate, mention share, per-model breakdown, and current-snapshot evidence. Same page you can view live for any firm on Atlas.
3
Identify gaps
Every prompt where a peer was cited and you weren't is a specific, fixable gap.
We cross-reference what the AI names you for against what it names competitors for. For each losing prompt we surface the specific competitor page that earned the citation and the pattern that made it citable — FAQ shape, schema markup, embedded quotes, specific numbers, statute references. AI assistants extract answers from patterns they trust; we tell you which pattern each competitor used and where to publish yours.
You see
A diff. Questions won. Questions lost. For each loss: the competitor page cited, the pattern to replicate, and the reason it was extracted.
4
Make changes
You publish the copy Viclaro spells out. Or your marketing team does. We hand you the shape.
For a firm, this usually means: one to three long-form FAQ pages hitting the exact buyer questions we found you missing on, JSON-LD schema on your existing service pages so answers get extracted verbatim, a "questions and answers" section on your homepage. Sometimes it is one paragraph on an existing page. Never a full site rebuild — the fixes are surgical because the gaps are specific.
You see
A checklist. Each item names the page to publish, the pattern to follow, and the specific buyer prompt it will move the needle on.
5
Rerun and compare
Same prompts, same models, same seeds, two weeks later. The delta is causal.
Every change in rank, share, or citation between runs is attributable to what changed on your pages between them — because everything else is held constant. No "well SEO takes six months to move", no "algorithm update", no confounds. If the rewrite worked, we can point at the prompt where it worked. If it did not, we can point at that too. Nobody gets to hide behind vibes.
You see
Before-and-after per prompt, per model, with the delta. Historical trend chart so you can watch a fix compound over multiple snapshots.