ChatGPT Ads Risk Misjudgment Under Click-Based Metrics
Maor Sadra argues that click-based attribution can make ChatGPT ads look weak by missing the conversations that shape purchases and send better-fit customers to sites.
Wire notes
- Sadra says marketers have spent 25 years optimizing around observable clicks.
- ChatGPT acquisition costs can exceed search costs, though several-minute conversations can qualify better-informed customers, according to Sadra.
- Sadra’s Hawaii example combines a friend’s wedding, a tight budget, their own room and surfing.
- Traditional last-click attribution can credit another channel when the buying influence began in an AI conversation.
- Sadra says early evidence looks different when marketers measure incremental business outcomes instead of last-click metrics.

Maor Sadra says performance marketers have spent 25 years optimizing for clicks, a habit that can make ChatGPT ads look weaker than they are.
In his Data-Driven Thinking column, Sadra rejects the standard test for a new advertising channel: whether it beats the last one on acquisition cost, conversion and overall performance. He says that comparison misreads conversational AI because search and conversational AI sit at different points in the customer journey.
What changes between search and conversational AI?
Sadra’s governing distinction is direct: “Search captures intent. Conversational AI shapes it.” According to him, Google built the world’s most successful advertising business by harvesting intent. A user searching for cheap hotels in Hawaii has already expressed commercial intent, leaving the advertiser to provide the most relevant answer when the purchase decision is made.
His example of conversational AI starts with more detail. The traveler explains that the trip is for a friend’s wedding, the budget is tight, the traveler wants their own room, and surfing matters too. Sadra says the exchange helps users compare alternatives and refine priorities before they reach a website to book.
Under Sadra’s model, the advertising opportunity extends beyond capturing demand at the final click. It includes participating in the conversation that develops that demand. Marketers need to evaluate ChatGPT on a different commercial basis, or they risk dismissing a significant channel before they understand how it creates value.
Does a higher CPA make ChatGPT ads a bad bet?
ChatGPT acquisition costs can appear higher than search costs, and Sadra says that was his own first reaction. His instinct was that a channel that looked and behaved like search should produce similar economics. He then contrasts a short keyword search with a user who spends several minutes discussing products, asking follow-up questions and refining requirements.
That longer interaction can qualify the customer before the visit to an advertiser’s website. In Sadra’s argument, the acquisition may cost more while delivering someone better informed and a stronger commercial fit. Sadra’s warning is direct: “Judging conversational AI purely on CPA risks confusing efficiency with effectiveness.”
In Sadra’s framework, a lower CPA can indicate efficiency without establishing effectiveness across the buying decision. Sadra says early evidence suggests a different picture when marketers measure incremental business outcomes instead of acquisition cost and last-click credit. On that basis, conversational AI can look very different from the view produced by last-click metrics.
Why do click-based metrics miss the conversation?
Sadra says performance marketing built its inherited frameworks around observable clicks for more than two decades. Attribution models rewarded whichever interaction came closest to conversion. That gave marketers a practical method for optimizing campaigns at scale in an internet dominated by search engines, display advertising and relatively linear customer journeys.
Conversational AI disrupts that sequence, according to Sadra. The interaction that influences a purchase can happen long before any click. If the eventual conversion comes later through another channel, traditional attribution naturally gives the credit elsewhere. The earlier AI exchange may still carry value; the framework simply lacks a way to capture it.
He describes the mismatch as an education gap. AI advertising can work effectively, but marketers can reach the wrong conclusion when they judge the channel through a system built to capture something else. Measuring incremental outcomes, he says, gives a more accurate view of the role conversational AI plays in buying decisions.
Should OpenAI ads be judged against Google?
Sadra frames the current debate around lower acquisition costs, conversion and the ability to win performance budgets, with Google Search as the benchmark. His preferred question is whether conversational AI should be judged like search at all.
He cites three earlier shifts in how marketers measure advertising: social required a model beyond display, retail media required a model beyond paid search inside a retailer’s website, and CTV required a model beyond linear TV metrics. Sadra presents ChatGPT advertising as another case that requires a new measurement mindset.
In Sadra’s model, Google earns from intent that already exists, while conversational AI has the potential to earn from the exchange that creates it. For marketers weighing ChatGPT ad budgets, his practical test is to examine incremental business outcomes and influence on the decision, not just the final click. Otherwise, he says, they may understate both the channel’s performance and conversational AI’s likely role in digital advertising.
via linkedin.com (Original)
More from James Calloway
Show full bio
Market editor covering technology and product news in marketing, advertising, marketing technology, media, SEO and search at Marketing Herald.
369 articles
More on the wire
- AEO Traffic Converts 3x-15x Better Than Traditional Search
- Brands Expect Little Holiday Sales Lift From ChatGPT Ads
- Zero Moment of Truth Expands From Three Sources to Five
- Google: AI Mode Queries Run Three Times Longer Than Traditional Search
- Attention, Not Data, Is Marketing's Real Currency, Says Joe Marchese