ChatGPT Ads and GEO: How Paid and Earned AI Visibility Work Separately — and Together

Go Fish Digital and OpenAI held an on-demand webinar published on 23 September 2026 examining how brands can pursue both paid and earned visibility inside ChatGPT — and why the two channels operate independently of each other.

Paid Ads and Organic Answers Are Separate Systems

Abhilash Edathil, who works on OpenAI's monetization team, demonstrated a travel-planning conversation in which a paid ad appeared alongside, but apart from, the organic answer. He stated that the ad placement can draw on the conversation's context and, where a user permits it, personalization, but was explicit that ads do not inform the answer itself.

AJ of Go Fish Digital distinguished between the two channels: a brand recommendation within an AI-generated answer represents earned visibility, while a clearly labeled advertisement in the same interface represents paid visibility. He noted that two different brands appearing in those respective positions is not a contradiction, and that a paid placement does not purchase inclusion in the model's answer.

Measuring GEO as a Pattern, Not a Fixed Ranking

AJ argued that a single prompt response is too variable to function like a traditional rank position. He proposed a three-part measurement framework: presence (whether a brand is mentioned or cited for a buyer question), representation (whether that description is accurate and current), and competitiveness (how often the brand appears relative to relevant rivals). He recommended testing roughly 20–40 questions that customers actually ask, running them repeatedly over a week or two under consistent model settings, and looking for recurring gaps rather than treating any single answer as a verdict.

Patrick Algrim of Go Fish Digital added that brands should examine the quality of a recommendation, not merely count mentions, because the evidence available about a brand's products and services shapes what the AI system can explain.

Website Evidence Must Come First

When asked whether to prioritize owned-site content or off-site mentions, both Go Fish speakers directed marketers to start with their own websites. Algrim used a moving-company example to illustrate the point: if a company offers cross-country moves but never states so on its site and has no corroborating evidence elsewhere, it should not assume an AI system will infer that service.

Algrim stated: "So start with your website because you can control it. Make sure everything is, you know, again, factually true, connected to a cohesive story about your brand, product, service, what it is that you offer, and, and really just start there."

AJ added a technical prerequisite: ensuring that relevant pages are accessible to search crawlers rather than blocked by robots rules or a CDN. He acknowledged that reviews, PR, and other external evidence still matter, but said they cannot rescue an unclear or contradictory account of the business on the brand's own pages.

Testing Ads and Interpreting AI Referral Data

Edathil said ChatGPT ads were shown to eligible adults in the Free and Go versions of ChatGPT, not paid versions, and that availability by market and vertical was still evolving at the time of the webinar. He directed advertisers to OpenAI's ads manager to check current eligibility and recommended defining the desired outcome — reach, traffic, or conversions — before running a test, and connecting conversion data through the available pixel or API.

AJ cautioned that AI influence can disappear from last-click reporting: a person may discover a brand in a chatbot, then search for it separately or paste a URL directly into a browser. He said GA4's AI-assistant referrals therefore show only part of the journey, and suggested supplementing that data with customer self-reporting and trends in branded and direct traffic. Algrim recommended testing specific website or campaign changes over time rather than attributing every traffic increase to a single source without evidence.