In an environment where AI retrieval is increasingly shaping how buyers, investors, journalists, and prospective hires form first-pass impressions, the audit most CMOs have not yet run is the one they need most.
Most CMOs have never asked ChatGPT what it thinks about their own brand. In a sample of 25-plus enterprise CMO conversations Everything-PR conducted across Q1 and Q2 2026, fewer than a third had run a structured AI engine prompt against their own brand. In an environment where AI retrieval is increasingly shaping how buyers, investors, journalists, and prospective hires form first-pass impressions, the gap is consequential.
Five minutes is enough to close most of it. Five questions. One ChatGPT tab. Run them this afternoon.
What follows is the diagnostic set. Each question is calibrated to expose a different layer of how AI engines have positioned your brand. The answers will surprise many readers — and across the conversations we've documented, the answers tend to shape how buyers approach the brand long before any sales conversation begins.
Question One What are the top three reasons not to buy [your brand]?
The prompt "What are the top three reasons NOT to buy [your brand]?"
This is the question that breaks the room.
CMOs flinch at it because the answer reads like an internal kill memo. AI engines tend to respond with three specific objections — price, quality, ethics, customer service, product limitations — sourced from criticism that appears with frequency across their training corpus. In many observed cases, those three objections track closely with the objections sales teams encounter in early-stage prospect conversations.
If you do not know what your three are, you cannot brief your sales team. You cannot adjust your messaging. You cannot prioritize your earned-media remediation.
Run it on three engines — ChatGPT, Claude, Perplexity. Compare. The engines often converge on the same three. Convergence across multiple engines tends to be the strongest available signal of a real reputational pattern.
Question TwoWho are [your brand]'s main competitors and how do they compare?
The prompt "Who are [your brand]'s main competitors and how do they compare?"
The AI engine has constructed a competitive frame for your category. Brands surfaced consistently across multiple engines tend to operate as the buyer's effective consideration set, regardless of internal market-share definitions.
This matters in two directions.
If an AI engine names a competitor you do not consider a competitor, the retrieval layer is collapsing your category in a way your strategic-planning team may not have modeled. Stanley was compared to Yeti across consumer search behavior for years before Stanley internally treated Yeti as a primary comparison. Retrieval frames eventually shape strategic frames.
If a competitor you obsess over does not appear in retrieval, the AI engine is signaling — though not proving — that buyers may not be comparing you to that competitor at the discovery stage.
Run this prompt across all five major engines. Watch which competitors are named in every answer versus only some. The universal-name set tends to reflect the real consideration frame more accurately than internal competitive lists.
Question ThreeWhat is [your brand] known for?
The prompt "What is [your brand] known for?"
This is the positioning audit.
You will receive a 50-to-150-word answer that summarizes what AI engines surface as your brand's identity. Compare that answer to the brand positioning document your agency built for you in your last brand refresh.
The two documents will rarely match.
The AI engine version reflects what your earned media, your owned content, your customer reviews, your Wikipedia entry, and your competitive coverage have collectively trained the retrieval layer to surface. The agency document reflects what you wished the brand stood for.
The agency version exists in slide decks. The AI version exists everywhere a research session begins.
If your agency positioning says you are the "innovative leader in [category]" and ChatGPT consistently describes you as the "established, traditional player," the buyer encounters the AI version before the agency version.
Question FourIs [your brand] a good company to work for?
The prompt "Is [your brand] a good company to work for?"
Most CMOs read this question and want to forward it to HR. Don't.
In an AI retrieval context, employer brand and consumer brand tend to be processed against overlapping source material. The buyer researching your product often runs this prompt too — not because they want a job, but because they are auditing your trustworthiness. The same applies to investors evaluating culture risk and journalists looking for the angle in their next piece.
The AI engine answer typically surfaces Glassdoor patterns, LinkedIn employee posts, Reddit threads, layoff news, and any union or workplace-controversy coverage. In many observed enterprise audits, the employer-brand answer shapes consumer-brand perception more than CMOs anticipate.
If the answer is unfavorable, your communications team has a project. If the answer is wrong — outdated, missing context, missing recent positive coverage — you have a GEO problem and a citation infrastructure investment to make.
Question FiveWhat is the most recent controversy involving [your brand]?
The prompt "What is the most recent controversy involving [your brand]?"
Every brand has a controversy in its history. The question is which one the AI engine surfaces first.
The answer will reveal the controversy with the longest citation tail. Bud Light's 2023 partnership decision continues to surface in AI prompts in 2026. Boeing's 737 MAX history continues to surface. Peloton's treadmill recall continues to surface. Citation tails tend to persist for years beyond the news cycle that produced them.
If the controversy surfaced is years old, your communications team has not yet displaced it through more recent coverage. If the controversy is recent, you are still inside the active crisis tail. If no controversy is surfaced, the question becomes whether the answer reflects clean reputation or whether you have a citation gap that more disciplined competitors will eventually exploit.
What to do with the five answers
The five answers form your AI Visibility Brief. Print them. Bring them to your next executive offsite. Ask three questions of every business unit leader and every brand manager:
The three offsite questions
Is this answer accurate?
Is this answer the answer we want?
What in our 2026 earned-media and content plan will move this answer in the direction we want by Q1 2027?
If the answers your team gives to those three questions are vague, your AI visibility strategy is not yet built. In observed cases across categories, buyers are not waiting for that strategy to mature. They are making consideration-stage decisions inside the answer that exists today.
The discipline this builds
Run the five questions quarterly. Track the answers over time. The drift in the answers — what we call Retrieval Drift — tends to be one of the most diagnostic available reads of brand health in 2026.
The five questions are not a one-time audit. They feed RLSOV — the master metric of AI-era brand visibility — and the cadence builds over quarters.
For the deeper twelve-prompt version of this audit, see The Prompts That Reveal Your Brand's Hidden Weaknesses. For the structured 30-minute version every CMO should run before their Q3 planning meeting, see The 30-Minute Self-Audit. For the explainer that defines every metric these questions surface, see How To Read A Retrieval Layer Share of Voice Report.
