Running an AI visibility competitive audit before a pitch
The most effective agency pitch in 2026 is to walk in with the prospect’s AI gap already mapped. Run a reverse-search competitive audit before the meeting: enter the prospect’s domain and their top competitors into the free Domain Check, then read the index backwards to see which buyer queries ChatGPT, Gemini and Grok cite each domain on. The deliverable is a side-by-side map — the high-intent questions where a competitor is named and the prospect is absent. That gap is concrete, undeniable, and screenshot-able, which is exactly what a cold prospect needs to take you seriously. The process: pick the right competitor set, run the checks, isolate the high-intent gaps, theme them into a story, and lead the pitch with the single most painful missing query.
Why audit before, not after
A generic pitch deck asks the prospect to imagine a problem. An audit shows them the problem, in their own category, with their own competitors. It flips the dynamic from “why should I hire you?” to “how fast can you fix this?” Because the raw data is free and fast to pull, there’s no reason to walk in empty-handed. This is the agency application of the same AI keyword gap mechanic the product is built on.
Step 1: choose the right competitor set
The audit is only as persuasive as the comparison. Pick three to five competitors the prospect actually loses deals to — not the biggest names in the space, but the ones in their consideration set. If you don’t know them, ask in the discovery call or infer from the prospect’s own “vs” pages and sales materials. The competitors that appear alongside the prospect in AI answers are themselves a strong hint about who the models treat as substitutes.
Step 2: run the Domain Check (step 1 of the audit proper)
The first concrete audit action is a reverse Domain Check. Run the free Domain Check on the prospect’s domain first, then on each competitor. For each, you get the queries that ChatGPT, Gemini and Grok cite the domain on, the intent behind each query, and which models named it. Capture the outputs side by side — this is your raw material.
What does the audit output look like?
The artifact you carry into the room is a short table of buying queries, scored for whether the prospect shows up and who beats them when they do not:
| Buying query | Prospect cited? | Competitor cited instead | Gap |
|---|---|---|---|
| best [category] tool for small teams | No | Competitor A | Yes |
| [category] software pricing compared | No | Competitor B | Yes |
| is [prospect brand] any good | Yes | — | No |
| alternatives to [competitor A] | No | Competitor A | Yes |
| how to choose a [category] provider | Mentioned only | Competitor B | Partial |
The brackets are placeholders — you fill them with the prospect’s real category and the competitor names the models actually return. The shape is what matters: it turns “we could help with AI visibility” into a specific, named gap.
Step 3: isolate the high-intent gaps
Don’t present everything. The persuasive subset is the high-intent queries where a competitor is cited and the prospect is not:
- “Best [category] for [use case]” — shortlist questions.
- “[Competitor] alternatives” — where the prospect should be named and isn’t.
- Comparison and “is X good for Y” queries — bottom-of-funnel evaluation moments.
Sort by intent so the questions closest to a buying decision rise to the top. A query a buyer asks right before they choose a vendor is worth far more than a generic informational one.
Step 4: theme the gaps into a story
A flat list of missing queries is data; a theme is a pitch. Look for the pattern:
- “You’re invisible on every comparison question in your category.”
- “Grok and Gemini name you, but ChatGPT — the assistant most of your buyers use — doesn’t.”
- “One specific competitor owns the AI answers for your highest-value use case.”
Model disagreement is its own story — the three assistants frequently cite different sources for the same question (see how often do ChatGPT, Gemini & Grok disagree on sources?). A prospect strong on one model and absent on another is a clear, fixable narrative.
Step 5: lead the pitch with the single worst gap
Open with one slide: the most painful high-intent query, the competitor named in the answer, and the prospect’s absence. Then widen to the theme and the plan. Don’t bury the lede in a 20-page report — the emotional impact is in the one undeniable gap.
Turning the audit into a contract
The audit is the wedge. Once the gap lands, the close is the same sequence covered in how to sell AI visibility: sell the paid audit/roadmap first, then convert to the recurring retainer add-on. Price it per how to price GEO / AEO services.
Try it on a real prospect right now — run the free Domain Check on their domain and one competitor, and you’ll have the spine of a pitch in minutes.