Three ways to audit. A 48-hour Snapshot to see whether there’s a real problem. A one-week Full Audit across four engines. Or Quarterly Tracking to stay honest over time. Every quote read by hand.
The audit work is the same in every tier — calibrated rubric, hand-read quotes, brand-variant matching, source tiering. What changes is breadth: how many prompts, how many engines, how often.
The smallest test we run. 10 buying-intent prompts on one engine, scored by hand, returned in 48 hours. Tells you whether there’s a real problem to fix — before you commit to the Full Audit.
100–300 buying-intent prompts. Four engines. Every quote read by hand, brand variants matched, sources tiered. You get four deliverables: the Score, the source map, the gap analysis, and a 90-day plan you start Monday.
Re-audit every quarter. Same prompt panel, same engines, same hand-reading. Track velocity, momentum, and volatility. Most clients catch a regression they would have missed alone — and a win they would have undersold.
Four deliverables, one report. A BeCited Score (0–100) with 95% confidence interval, computed from five signals across four engines. A Source Influence Map of every platform AI cites in your category. Every prompt gap ranked by revenue impact, with a root cause attached. A 90-day plan, prioritized per engine.
What works on Gemini doesn’t work on ChatGPT — we tell you exactly what to do for each. Every gap gets a root cause so you fix the real problem, not the symptom.
See the Methodology →A single “visibility” number hides everything that matters. The audit measures eight distinct primitives so you can see where you’re weak and why.
How often each of the four engines mentions your brand, by category and by intent tier. Lets you see if you’re invisible everywhere or just on one engine.
When you’re mentioned, are you actually being recommended — or just listed in passing? Mention isn’t recommendation, and the audit separates them.
Being the first-listed recommendation matters more than being fifth. Position 1 gets a 1.25× multiplier, position 3+ gets 0.85× — reflects how users actually read AI answers.
Your slice of the engine’s recommendation pie versus your real competitors. Anchored to who you actually win and lose against, not to a generic competitor list.
Which platforms AI cites in your category, and how those break down between primary, secondary, and tertiary sources. Tells you where to spend remediation effort.
Every gap gets a root cause: not on a key source, competitor dominance, content gap, review deficit, category mismatch. Answers why you’re absent, not just where.
In Quarterly Tracking, every action you took since the prior audit is correlated with the dimension changes that followed. You see what worked, not just what changed.
Across three or more audits we compute velocity, momentum, volatility, and a forecast. Lets you spot a regression before it becomes a quarter of lost pipeline.
If you’ve never audited AI visibility before, start with the Snapshot. It’s a 48-hour gut check on whether there’s a real problem to fix — and it credits 100% toward a Full Audit if you upgrade within 30 days. If you already know AI search is hurting you and you need a plan, go straight to the Full Audit.
Breadth and engines. A Snapshot is 10 prompts on one engine — enough to show whether AI knows you exist in your category. The Full Audit is 100–300 prompts across all four engines with brand-variant matching, source tiering, root-cause attribution, and a 90-day plan. The methodology is the same; what scales is coverage.
By hand. AI engines drift — “Acme Co.” in one answer is “Acme Cleaning” in the next, and a script counts them twice. Citations are sometimes fabricated outright. Every quote gets read once by the script and once by Owen, brand variants get matched back to one company, and fake URLs get flagged. See the methodology for the full process.
That’s what Quarterly Tracking is for. Same prompt panel, same engines, same rubric — re-run every 90 days. Adds velocity, momentum, volatility, and a forecast on top of every dimension, plus action-to-outcome attribution if you tell us what changed between audits.
Yes — that’s a separate engagement called Relevance Engineering. The audit gives you the plan; relevance engineering executes it. Most clients do the audit first, then decide which fixes to outsource and which to keep in-house.
Send a quick note. I’ll reply with a recommendation, a price, and a kickoff window — usually within a day. No sales call required.
Contact Owen →