Brands Show Up in ChatGPT, but 56% of Marketers Can't Prove ROI

Fifty-six percent of small and medium business marketers say measuring ROI from AI visibility is their top challenge, according to an Emarketer survey. That single number explains most of what is happening in the newest corner of the search industry. Brands want to appear in ChatGPT answers. They cannot yet prove it pays.

ADWEEK's Tech Advantage coverage names the companies pushing forward anyway: Flo Health, Motorway, Glance, and Business Intelligence Group are all working to improve their visibility inside AI-generated answers. The report is blunt about the structural problem, though. The industry, it says, is hobbled by a lack of standardization. No shared metric exists. No shared methodology either.

How big is the demand?

The numbers point one direction. In a survey of 75 small and medium business search marketers conducted by Emarketer together with Semify, 76% said appearing in AI search is extremely or very important to them. Roughly three in four respondents, in other words, treat chatbot mentions as a priority channel, not an experiment.

Demand that size builds a market fast. Getting recommended by an AI chatbot, as ADWEEK puts it, now has its own burgeoning cottage industry.

Who is selling the solution?

Two kinds of vendors have moved in.

• Agencies with proprietary tooling. Jellyfish, for one, has built in-house tools to chart where and how brands appear in answer engines such as ChatGPT.
• Startups selling software. A wave of new companies offers platforms designed to understand and improve a brand's AI visibility.

ADWEEK profiled 11 of those startups. Profound stands out among them, and not for its technology claims alone. It is among the best-funded of the group, having raised $180 million in a Series D round at a $1.8 billion valuation, according to the report.

That valuation deserves a pause. Investors are pricing this category as if the measurement problem were already solved. The survey data says it is not.

What does the 56% number actually mean?

Read the two figures side by side and the tension is obvious. Seventy-six percent of surveyed marketers say AI search visibility matters enormously. Fifty-six percent of SMBs, per Emarketer's survey, cite measuring the ROI of that visibility as their biggest obstacle. Marketers are committing budget to a channel whose returns they cannot reliably count.

This is not an unusual pattern in ad tech. It resembles the early years of social media attribution, when spending preceded measurement by half a decade. The difference here is speed. The tooling gold rush and the measurement gap opened at the same time, and vendors are selling dashboards for a metric nobody has agreed to define.

The standardization gap has practical consequences. Without common definitions, one vendor's "visibility score" cannot be compared with another's. A brand that climbs in one tool's rankings may hold flat in a competitor's. Procurement teams, and CFOs, will eventually ask which number is real.

What did the brands actually achieve?

Here is what the coverage does not say, and it matters. ADWEEK names Flo Health, Motorway, Glance, and Business Intelligence Group as firms improving their AI visibility. The article does not, in the accessible portion, attach revenue figures, conversion lifts, or traffic deltas to any of them.

That silence is the story. Case studies without outcome numbers are the classic signature of a channel in its pre-proof stage. The names signal seriousness. The absence of attached metrics signals that nobody is ready to publish the scoreboard yet.

The $180 million raised by Profound says investors will accept that ambiguity for now. A $1.8 billion valuation on a startup whose core product measures a non-standardized metric is a bet that standardization will eventually consolidate around whoever scales first. That can work. It has also burned investors before in adjacent categories.

What should marketing leaders do?

Treat AI visibility as real but unproven. The 75-respondent survey base is small, so read the percentages as directional rather than definitive. If 76% of your peers consider the channel important, ignoring it carries its own risk. If 56% cannot measure returns, budget accordingly.

The practical posture for a CMO is straightforward. Track appearances in answer engines, using agency tools like Jellyfish's or startup platforms. Demand that vendors explain their methodology in writing. And hold off on declaring AI visibility a performance channel until someone, vendor or industry body, publishes a standard that two competing tools can agree on.

The measurement gap will close eventually. Between now and then, the smart money stays curious and skeptical at the same time.