The Five Questions to Ask Any Agency Pitching You AI Readiness

The Five Questions to Ask Any Agency Pitching You AI Readiness

Every week, another agency launches an “AI optimization” service. They promise to get you mentioned in ChatGPT, boost your “AI rankings,” and protect your marketing for the long term. Most of these services are repackaged SEO with a new label. Some are actively harmful.

Ask these five questions. The answers will tell you whether the agency understands AI readiness as a system or is selling you a feature dressed up as a strategy.

Question 1: AI readiness has three layers, not one

Ask the agency to describe how AI readiness works. If their answer focuses on one thing, usually content creation or schema markup, they do not understand the system.

AI readiness has three layers. Clarity is whether machines can read and trust your data. Coverage is whether machines choose you when buyers ask relevant questions. Conversion is whether the visitors machines send you actually convert into pipeline.

These layers are sequential and constraining. Clarity caps Coverage. If machines cannot understand your content accurately, they will not recommend you, regardless of how much content you produce. Coverage caps Conversion. If machines send you the wrong visitors or set incorrect expectations, your conversion rate suffers.

Most “AI readiness” packages work the same way. The agency restructures a few pages with clearer headings. They add schema markup. They create an FAQ section. Maybe they publish AI-written content targeting long-tail queries. None of this is worthless. Schema markup helps machines read your data. Specific language is better than vague language. FAQ content can match real buyer queries. But calling this “AI optimization” is like calling a coat of paint a structural renovation. The site looks more organized. The machines still cannot trust your data, your visitors still cannot find the proof they need, and your conversion rate on AI-referred traffic stays where it was.

Any agency that talks about only one of these layers is not solving the problem. They are optimizing one input and hoping the rest works out. It will not. The ceiling rule means that if you fix Coverage without fixing Clarity first, you hit a hard ceiling on results. The upstream problem limits the downstream outcome.

A real operator will describe the system in layers, explain how they relate, and start with the layer that is constraining everything else.

Question 2: Measurement methodology separates real operators from guessers

Ask how they measure whether your AI readiness is improving. If they talk about “rankings” in AI tools or show you screenshots of ChatGPT answers, they are guessing.

AI readiness requires deterministic measurement. That means running structured queries across multiple AI platforms, documenting whether your company appears, what the AI says about you, and whether the claims are accurate. Then repeating that measurement over time to track changes.

The key metric is inclusion rate. What percentage of relevant buyer queries include your company in the AI’s answer? Not where you appear in a list. Whether you appear at all. This is a binary measure with a quality dimension: you either appear or you do not, and when you appear, the AI either cites you with confidence or hedges.

Probabilistic approaches, like sampling a few queries and making qualitative observations, are not measurement. They are anecdote. An agency that cannot tell you your inclusion rate across a defined set of buyer topics is not measuring anything meaningful.

A real operator will explain their query methodology, show you the topic list they track, define how they score inclusion, and demonstrate how they measure change over time.

Question 3: A real Clarity diagnostic covers eight specific factors

Ask what their diagnostic measures. If they cannot name the factors, they are not running a diagnostic. They are doing a content audit with a new name.

A real Clarity diagnostic measures eight weighted factors: Context (20%, whether machines can determine what you do and who you serve), Accuracy (17%, whether your claims match what external sources can verify), Consistency (15%, whether you say the same things across all public surfaces), Specificity (15%, whether your claims are precise or vague), Recency (12%, whether your content is current and regularly updated), Machine Readability (8%, whether your technical structure supports extraction), Rendering Independence (6%, whether your content renders without JavaScript so all crawlers can access it), and Hallucination Risk (7%, whether gaps in your content cause AI to fabricate information about you).

These are measurable, improvable factors. Each one can be scored, benchmarked, and tracked over time. An agency that lumps all of this into “content quality” or “technical SEO” is not doing the work. They are applying old frameworks to a new problem.

Most agencies skip the cross-platform data audit entirely. Your Google Business Profile lists a different address than your website. LinkedIn shows a different company description than your sales deck. Your distributor’s site has outdated pricing while your site shows current numbers. Each contradiction teaches AI systems to distrust your data. No amount of on-page schema fixes that. Your content is probably written for Explorers who need category education, not Validators who arrive pre-educated and need to confirm specific claims in seconds. The mismatch between what the visitor expects and what the site delivers is where most AI-referred traffic dies.

A real operator will score each factor, show you where you lose points, and map specific improvements to specific factors.

Question 4: Upstream and downstream are different disciplines

Ask whether their team handles Clarity and Coverage with the same people who handle Conversion. If yes, they are generalists pretending to be specialists.

Clarity and Coverage are upstream disciplines. They involve making your data machine-readable, structuring your content for extraction, aligning your topics with buyer research patterns, and building evidence strength so machines cite you with confidence. This is data architecture, content strategy, and technical structuring work. It requires understanding how AI systems parse, correlate, and synthesize information.

Conversion is a downstream discipline. It involves redesigning pages for Validator behavior, building proof into the experience, reducing friction for pre-educated visitors, and structuring decision paths that confirm rather than persuade. This is UX, copywriting, and conversion optimization work.

These require different skillsets. The person who restructures your content for machine extraction is not the same person who redesigns your pricing page for Validator conversion. An agency that blurs these disciplines is either weak at one of them or treating both superficially.

A real operator will explain which discipline they specialize in, how they handle the other, and where they draw the line between upstream and downstream work.

Question 5: AI hallucinations are a measurable, persistent problem

Ask how they track what AI gets wrong about your company. If they do not have an answer, they are ignoring the most dangerous part of AI readiness.

AI hallucinations about your company are not rare edge cases. They happen regularly. The machine fills gaps in your content with plausible-sounding fabrications. It attributes capabilities you do not have, cites partnerships that do not exist, or misrepresents your pricing. These hallucinations are persistent. Once embedded, they compound as other AI systems learn from the same incorrect data.

If an agency does not measure hallucination risk as part of their diagnostic, they cannot protect you from it. They will improve your visibility while making the accuracy of that visibility worse. More AI mentions of your company, but a growing percentage of those mentions contain false claims. That is not a win. That is a liability.

Hallucination tracking requires running queries about your company across AI platforms, documenting what each platform says, flagging inaccuracies, and measuring whether corrections to your content actually fix the hallucinations over time.

A real operator will show you a hallucination report. They will know what AI gets wrong about you, how often, and in which platforms.

The real test is whether they understand AI readiness as a system

These five questions test for one thing: whether the agency understands AI readiness as a system with interacting parts, or is selling a point solution for a single symptom.

Part of the problem is incentives. Most agencies are paid to produce deliverables, not outcomes. Restructuring pages and adding schema produces visible deliverables the client can see and the agency can bill for. Running a cross-platform data consistency audit, scoring evidence strength across your entire site, and building a conversion system that adapts to visitor intent is harder to scope, harder to price, and harder to sell. The “AI optimization” package becomes a repackaged version of the content work the agency was already doing, with a new label that matches the current moment. The client feels covered. The agency keeps the retainer. Neither notices that the actual problem went untouched.

The agencies worth working with will describe a system. They will talk about layers, ceiling effects, compounding loops, and the relationship between upstream inputs and downstream outcomes. They will measure things precisely. They will specialize in the part of the system where they are strongest and be honest about where they need to bring in complementary expertise.

The agencies to avoid will sell you a feature. “We get you mentioned in ChatGPT.” “We restructure your schema markup.” “We write AI-friendly content.” Each of these can be useful in isolation, but none of them is a strategy. A point solution applied to a system problem produces point results. A small improvement in one area that hits a ceiling because the constraining layer was never addressed.

The ceiling rule is the key mental model. Clarity caps Coverage. Coverage caps Conversion. The layer that is most broken is the one that limits everything else. Fix that layer first, then move to the next. That is how you build AI readiness on the C3 Framework, not an AI optimization project.

Any agency that cannot articulate that system is not ready to help you build it.