OpenAI’s Ad Bet Needs Google-Style Intent Data to Scale

Google generated nearly $295 billion in advertising revenue last year. Search and other properties produced $225 billion, about $4.3 billion a week—more than The Trade Desk, the largest independent demand-side platform, booked in a full year at $2.9 billion.

That gap frames Patrick Briggs’s argument. The Semify CEO begins with a blunt proposition: “OpenAI wants to be Google.” He expects OpenAI to reject the public comparison, but says the advertising economics make it unavoidable. He calls Google’s ad operation one of the best businesses in history because it places companies in front of people when they state a need.

Briggs says ChatGPT now reaches that same moment of expressed intent. The comparison with Instagram matters. People may open Instagram to browse without needing anything; they open Google for a specific blue sweater.

In ChatGPT, many users go further. Instead of typing “sweater,” someone can explain that they need warmth for a week in Chicago in January, that wool irritates their skin, and that they do not want to spend more than $150. The exchange becomes a brief containing wishes, constraints, exclusions, and concerns.

Then comes the decision. The shopper compares options and chooses a product, which is the point at which a sweater brand wants to appear. Briggs says OpenAI needs analytics that reveal consumer intent based on those prompts.

What can advertisers already buy?

The transaction layer already exists. OpenAI’s beta self-service Ads Manager lets advertisers buy on a CPM or CPC basis, use a pixel and Conversions API to measure conversions, and view campaign-level performance reports. OpenAI also tests Sponsored Agents, which opens a separate conversation with an advertiser after a user clicks an ad.

What those tools do not yet provide is the context advertisers get from Google Search. A search advertiser chooses keywords, reviews the searches that triggered ads, identifies which searches became customers, and uses negative keywords to block wasteful terms. A ChatGPT ad may appear beside a conversation, but its buyer cannot yet see the conversation’s intent.

Google has offered some version of that demand map for two decades. Briggs wants equivalent reports for ChatGPT: the categories of intent behind impressions, whether users were exploring, comparing, or ready to act, and which needs or product criteria shaped their engagement. He also wants marketers to know which sweater prompts produced business and which did not.

Can intent data remain private?

Briggs accepts that ChatGPT conversations may contain information people would never enter into a search box. His answer is aggregation, not individual access. No advertiser should read a conversation or see an individual prompt, he argues. Google already limits, aggregates, and buckets queries too rare to disclose while preserving enough information for campaigns.

The advertising industry also converts web behavior into pseudonymized, aggregated signals about audiences and intent. OpenAI says advertisers receive aggregated performance insights without access to individual chats. Briggs proposes applying the same protection to commercial intent: classify conversations by needs and decision stage, then report patterns without unlocking the underlying chats.

What would change advertising budgets?

A chief marketing officer decides which platform can prove business value. If ChatGPT offers only reach, impressions, and clicks, Briggs expects it to sit in the experimental or awareness category, drawing from innovation budgets or paid social. If it links privacy-safe intent categories to business outcomes, it can compete for paid search budgets meant to capture existing demand.

Marketers can test that distinction now, Briggs says. They can run ChatGPT against paid search and paid social, ask OpenAI representatives what intent reporting exists, and keep asking as the product develops. Until OpenAI exposes the intent behind clicks, he says, advertisers should not pay search prices for it.

The underlying advantage is already in place, in Briggs’s view. ChatGPT users describe goals, problems, preferences, objections, and purchase criteria in their own words. The missing product is not another ad placement; it is an advertiser’s map connecting those expressions to results.

Better reporting could also improve the ad itself. A user asking for an itch-free sweater below $150 should see one. That is both the measurable business case and a useful experience for the shopper.

Briggs’s prescription for the next phase is exact: “It has to make that intent measurable without making it personal.”