OpenAI and Go Fish Digital say a ChatGPT ad does not buy a mention in the answer

A brand paying for a ChatGPT ad slot next to a shopping conversation does not buy a place in the model's actual recommendation. Search Engine Journal reported on 23 September that in a joint webinar hosted by SEJ, OpenAI's monetization team and the agency Go Fish Digital walked marketers through how paid placements and earned answers work as separate slots, and what to measure now that the same conversation contains both.

The Search Engine Journal report names the speakers. Abhilash Edathil of OpenAI's monetization team showed a travel-planning conversation in which an ad appeared alongside, but apart from, the organic answer, and said the placement can draw on the conversation's context and, where the user permits it, personalization. He was explicit that ads do not inform the answer. AJ of Go Fish Digital called the answer's brand recommendation earned visibility and a labeled ad paid visibility, and said two different brands appearing in those positions is not a contradiction.

Measure GEO as a pattern rather than a rank

According to the SEJ report, AJ argued that a single prompt response is too variable to function like a traditional rank position, so treating one answer as a verdict is a mistake. He proposed three measurable things: presence (is the brand mentioned or cited for a buyer question), representation (is the description of the brand accurate and current), and competitiveness (how often does the brand appear against relevant rivals).

To make that assessment useful, AJ recommended starting with roughly 20 to 40 questions that customers actually ask, running them repeatedly over one or two weeks, and keeping the model and settings consistent. Look for recurring gaps rather than reading one answer as a result. Patrick Algrim of Go Fish Digital added that quality matters as much as count: the evidence a model can find about a company's products and services shapes what the system is able to explain about them.

Fix the evidence on your own site before chasing mentions

Both Go Fish speakers told SEJ that the first place to work is the company's own website. Algrim used a moving-company example: if a firm offers cross-country moves but never says so on its site and has no corroborating evidence elsewhere, it should not expect an AI system to infer that service. Own pages need to answer the questions a buyer would ask, name what the company actually does, and stay consistent with external descriptions.

AJ added a technical first check: confirm the relevant pages are readable by search crawlers, and not blocked by robots rules or a CDN. Reviews, PR, and other external evidence still count, but they cannot rescue an unclear or contradictory account of the business.

On measurement, Edathil said ChatGPT ads were shown to eligible adults on the Free and Go tiers only, and that availability by market and by vertical was still changing at the time of the webinar. He described those conditions as of the recording, so any advertiser should check current eligibility in OpenAI's ads manager. AJ warned that GA4's AI-assistant referrals show only part of the buyer journey, because a person may find a brand in a chatbot, then search or paste a URL into a browser, so branded search trends and customer self-reporting have to close the gap.

Source

• ChatGPT Ads And GEO: Where Paid And Earned AI Visibility Fit Together, Heather Campbell, Search Engine Journal, 23 September 2026