AI visibility / The 12AM method
Our AI visibility method records the questions, answers, and sources behind a finding. See what we measure, how we compare observations, and where the evidence stops.
01 / Define the observation
Known vs. chosenWe distinguish four descriptive levels in the questions we review. They are observations about an answer, not promised stages of growth.
Ask by name and AI cannot reliably identify who you are.
Not reliably identifiedAsk by name and AI answers — but it does not volunteer your firm.
Identified when asked by nameAI brings your firm up when the buyer never mentioned your name.
Mentioned without a branded promptAI actively puts your firm forward as a firm to consider.
Suggested as a firm to considerKeep question wording, platform, date, sample size, and counting rules with each comparison. Inquiries and signed matters require separate analytics and intake records.
02 / Follow the source families
An explanation of the work beneath the answerYour website is one source. Profiles, reviews, legal directories, and independent coverage add context. We examine how that record represents your firm.
“Who handles
complex estate
planning in Dallas?”
Service pages, attorney bios, and client questions establish the record your firm publishes.
These six source families contain the 16 signals 12AM reviews in the Diagnostic. This is an illustrative pathway — not an engine trace, ranking formula, or live audit.
03 / Inspect the 16 signals
16 signals · proprietary weighting model12AM reviews sixteen signals across three evidence layers: what you publish, the profiles you control, and what independent sources say about you. We publish what we evaluate; the weighting model and scoring logic remain proprietary to 12AM.
Bigger does not automatically mean more visible to AI. Large firms often have more evidence because scale creates more reviews, mentions and coverage. But evidence can be built deliberately.
Signal-density bands in the Diagnostic are descriptive observations, not proven thresholds. Scoring weights are 12AM's stated method.
04 / The measurement register
These are complementary ways to investigate the record, not seven sequential steps. The report identifies the questions, platforms, dates, and scope behind its findings.
Establish the starting view of your firm and competitors.
Record the scope and baseline before comparing later runs.Repeat an agreed set of buying questions.
Preserve the wording. Record the platform and observation date.Investigate evidence beyond the tracking-panel results.
Distinguish supporting sources, missing evidence, and assumptions.Inspect pages surfaced on commercially relevant questions.
Keep source URLs with the findings they support.Compare responses across ChatGPT, Gemini, Claude, and Perplexity.
Separate mentions, citations, and recommendations in the recorded answers.Examine backlink, content, and organic-search records.
Keep the underlying exports and observation window available.Check crawler access, page markup, profiles, and credentials.
Verify the source page or credential issuer directly.A report must state its own sample size. A prompt count or page count from one engagement does not define every Diagnostic.
05 / What the Diagnostic gives you
The paid AI Visibility DiagnosticIf the free scan shows a meaningful gap, the paid Diagnostic establishes the full baseline: your firm, your market, the evidence behind the gap, and a plan for what to do next.
If the free scan shows a meaningful gap, we confirm scope, pricing, and any implementation credit before the Diagnostic begins.
Start With the Free ScanWe confirm practice areas, markets, question set, and delivery timing before work begins. A written report and a strategist readout, yours to keep.
Questions about the engagement?
/ INTELLIGENCEActual report findingsThe recommendation gap
One firm. Fifty buying questions.
A measurable gap in who AI names.
See exactly where you’re missing.Then inspect the competitors and questions behind the gap.
Findings from the redacted sample. These counts can overlap; they are not separate shares of one total.
/ ACTIONActual report structureYour 90-day roadmap
Work sequenced by dependency.
Progress measured against the same baseline.
Technical access, identity and core pages.
Profiles, directories and submissions.
Pages and answers tied to the findings.
Video, newsletters and expert commentary.
Rerun the questions. Compare the evidence.
Delivery clockThe work we control.
Response clockWhen AI systems reflect it.
Summarized from the sample calendar. Client scope varies. AI response timing and outcomes are not guaranteed.
Visual summaries of actual redacted report pages. Open the originals to inspect the evidence. The proprietary weighting model stays inside the client Diagnostic.
Use the roadmap with your team, your current agency, or 12AM.
06 / Trace a finding
A visibility claim should lead back to a question and a recorded observation. Here is one buyer scenario from our redacted Diagnostic sample.
Redacted Diagnostic · The Gap“Twelve rental units through a patchwork of LLCs, and a new lawsuit exposed gaps”
Comparison figures as printed in the sample report
The report records a naming gap for this buyer scenario. The comparison does not establish an endorsement, explain the cause, or predict future answers.
This report extract summarizes findings. It does not include engine-level answer transcripts or citation URLs. Those details cannot be independently checked from this extract alone.
07 / Compare carefully
AI answers can change with wording, platform, location, and date. We repeat the procedure and document changes; we do not promise identical answers.
Use the same agreed questions for follow-up comparisons. Record any changes in platform, scope, or collection conditions so readers can judge comparability.
The scoring weights reflect 12AM’s method. They remain proprietary. They do not predict recommendations, leads, or revenue.
A source link, a mention, and a recommendation describe different observations. Business outcomes require separate analytics and intake evidence.
A directory listing or article adds to the public record. Its presence does not establish that an AI system will cite or recommend your firm.
Your report. Your decision.
You keep your Diagnostic and roadmap. Share them with your team, your existing agency, or 12AM.
12AM retains its scoring model and underlying research. We confirm deliverables, access, and responsibilities before the work starts.
See how we implement the roadmapYour firm reviews facts and professional claims before publication. The firm remains responsible for applicable attorney-advertising requirements.
08 / Bring these questions to the review
Use the same questions when you evaluate 12AM or another agency.
Explore the research archiveStart with your firm
Share your firm details, then book a 30-minute review of your Free AI Visibility Scan. Decide on the paid Diagnostic after the review.
