Barnard says AI visibility needs a new measurement model

The old dashboard breaks in AI search

AI visibility has outgrown rank reports. Jason Barnard’s core argument is that the question agencies keep asking, how to measure presence in ChatGPT, Perplexity, Google’s AI Mode, and other AI surfaces, cannot be answered with the same tools that once tracked blue-link rankings.

That matters because the old search model assumed a finite results page, stable positions, and a click you could count. AI systems do something messier: they synthesize, cite, surface follow-up prompts, and sometimes push the user straight into an agentic workflow. In that environment, a keyword list and a shinier rank tracker do not solve the real problem. They only make the report look cleaner than the data deserves.

Why old SEO math keeps failing

Barnard is pushing back on a familiar agency habit: taking a broken measurement problem and dressing it up as a tooling problem. Traditional SEO measurement was built around page rank, keyword positions, click-through rate, and a limited set of result pages. That model made sense when search behaved like a shelf with labeled products. AI search behaves more like a conversation that can branch, summarize, and act.

The shift is not theoretical. Google began rolling out AI Overviews to everyone in the U.S. on May 14, 2024, after first announcing Search Generative Experience in May 2023. OpenAI launched ChatGPT search on October 31, 2024, then expanded it to more ChatGPT users in December 2024. Perplexity has positioned itself as a free AI-powered answer engine that gives accurate, trusted, real-time answers. By the time Google I/O 2026 opened on May 19 and 20, AI-driven search was no longer a side experiment. It was central to the product roadmap.

That is why Barnard calls AI visibility a macro measurement problem, not a single-KPI problem. If the interface itself now blends answer, citation, and action, then any one number will flatten the story.

What the Funnel Query Pathway changes

Barnard’s answer is the Funnel Query Pathway, which he treats as strategy, measurement, and analysis at the same time. The value of that framing is that it follows the user’s journey instead of pretending every query deserves the same report row.

For agencies, that means tracking visibility across the stages that actually matter to business outcomes:

• Discovery queries, where a brand needs to show up in broad, problem-seeking prompts.
• Consideration queries, where the engine is comparing options, citing sources, or recommending specific names.
• Conversion-oriented queries, where the user is closer to action, whether that means a visit, a signup, a purchase, or a next-step prompt inside the answer engine itself.

This is the part of the conversation where clients need honesty, not theater. Barnard’s point is that agencies should not sell a fake precision story, because there is no clean dashboard that reliably shows presence, visibility, and action all at once. The better move is to document how a brand appears across that funnel, then connect those appearances to evidence of influence.

Presence is not the same as performance

One of the most important changes in AI search is that “being found” no longer means the same thing it did in classic SEO. A brand can be present in an answer engine without earning a traditional click, and it can influence a decision even when the interaction never looks like a normal search session. Google’s AI Overviews made that tension obvious by pairing synthesis with links to web sources, which creates visibility without always creating the old kind of measurable visit.

That is why agency reporting needs to move beyond a single rank or traffic metric. A useful measurement model should capture whether the brand is being surfaced in the right questions, whether the engine is citing or recommending it, and whether those appearances appear to shape the next step. In practical terms, the reporting stack needs to answer three questions: Are we in the conversation, are we part of the answer, and does that exposure influence action?

Barnard’s framework gives leadership teams a way to explain why the old reporting model is now misleading. The market does not need more confidence theater. It needs an operating model that can handle answer engines, assistants, and agents without pretending they work like a standard search results page.

How agencies should talk about AI visibility now

The best agencies will treat this like a measurement design problem, not a tooling contest. That means building custom reporting around the queries clients actually care about, then separating discovery, consideration, and conversion evidence instead of collapsing all of it into one score. It also means documenting where visibility is appearing, because AI search now spans multiple surfaces rather than one canonical page.

A practical framework should include:

• The query set, broken into awareness, comparison, and action intent.
• The surface, whether it is ChatGPT, Perplexity, Google’s AI Overviews, AI Mode, or another answer environment.
• The type of presence, such as citation, mention, recommendation, or inclusion in a synthesized answer.
• The downstream effect, including assisted visits, branded demand, lead quality, or conversion behavior tied back to that exposure.

That approach does not promise certainty where none exists. It gives clients something better: a measurement model that reflects how people actually search now.

Why this is also an agency value story

Barnard’s bigger warning is commercial as much as it is technical. As the market shifts away from simple ranking-based measurement, agencies that still sell a neat keyword report will look increasingly out of touch. Agencies that can define and document a new model will be able to set expectations clearly, build custom reporting, and defend their value when clients ask why AI visibility looks different from classic SEO.

That is the real accounting challenge in AI search. Not how to force a new medium into an old dashboard, but how to prove influence when visibility is spread across discovery, consideration, and conversion, and when the answer itself may be the product. The agencies that get that right will not just report better. They will sound like they understand the market they are being paid to navigate.