The Shift Underway: From Clicks to Answers

Advertising inside AI assistants exists for one reason: search behavior is moving toward AI-generated answers, and the “answer surface” is becoming where attention concentrates. The evidence is concrete and dated. Gartner predicted in February 2024 that traditional search-engine volume would fall 25% by 2026 as people shift to AI chatbots and virtual agents. SparkToro measured US zero-click searches — the ones where the answer is delivered in place and the user never clicks out to a website — rising from 58.5% in 2024 to 68.01% across the first four months of 2026.

The strategic consequence is direct. As fewer queries end in a click to a website, the AI answer itself becomes the surface that captures intent — which is exactly the bet ChatGPT Ads, Google AI Overviews, Microsoft Copilot, and Amazon Rufus are each making. That is the honest starting point of this page: the direction is real and measurable, even where the size of the in-chat ad opportunity is contested. For what these ads actually are, start with what ChatGPT ads are; for the full reach and forecast data set and the head-to-head against Google and social, see ChatGPT ads vs Google and social ads. The click-versus-answer measurement consequences belong with measuring ChatGPT ads performance.

MB Adv Agency runs the paid-media programs — Google Ads, Meta, and the broader paid mix — that a new channel like ChatGPT Ads slots into, so our team treats “where do answers get delivered” as a first-order question rather than a curiosity. When the surface that answers a query changes, the whole search economy the businesses we work with depend on changes with it.

Key Takeaways

• The shift is real; the in-chat ad market is contested. Gartner projected traditional search volume down 25% by 2026 and SparkToro measured US zero-click searches at 68% in early 2026 — but how much advertising lands inside chatbots specifically is where forecasts diverge sharply.
• eMarketer’s realistic read: chatbots reach only ~$5B by 2030. It forecasts $68.25B in total US AI ad spend by 2030, but says 80%+ is “AI-adjacent” and standalone chatbots earn only about $5B of that.
• The ~$100B-by-2030 figure is an outlier, never consensus. It is OpenAI’s own internal projection (reported via Axios and The Information) and a matching Barclays model — an aggressive ambition, not a market forecast.
• The bets are diverging, not collapsing. Perplexity launched ads and then publicly dropped them on February 18, 2026 over trust, while OpenAI, Google, Microsoft, and Amazon all keep monetizing their AI surfaces — different companies choosing differently on the same shift.
• Prepare with a test, not a bet. Run small measurable tests, watch the trust signals, build first-party data, and treat GEO/AEO as the complementary organic track to paid AI placement.

GEO and AEO: The Discipline That Pairs With Advertising in AI

In an answer-first web, brands compete on two tracks — and only one of them is advertising. The organic track is GEO (Generative Engine Optimization), being cited or recommended inside AI-generated answers such as ChatGPT, Perplexity, Gemini, and AI Overviews, and its close cousin AEO (Answer Engine Optimization), being the extracted answer in featured snippets, voice results, and AI answer boxes. The paid track is ChatGPT Ads and its peers: buying the sponsored placement outright. Both are industry coinages, not official OpenAI or Google terms.

The point that matters for a forward-looking plan is that the two are complementary, not interchangeable. You compete to be the source the AI cites and, separately, to buy the sponsored slot — and the strongest programs think about both. GEO and AEO build on the same foundation as SEO: crawlable, authoritative, well-structured content. They simply optimize for a different outcome — being named in an AI answer rather than ranking a blue link. The honest caveat is that GEO is early: its metrics and best practices are immature in 2026, so it is a developing discipline to invest in, not a solved one.

This pillar keeps the paid depth in-cluster and routes the organic depth to its sibling glossary. For the full mechanics of the organic side of the same shift, see how AI is changing SEO. And because AI assistants increasingly blend answers with commerce, note that ChatGPT’s own product discovery and checkout — the shopping and agentic-commerce layer — is organic, not advertising; OpenAI states retailers cannot pay for higher placement there.

Different Bets on the Same Shift: ChatGPT, Perplexity, Google, Microsoft, Amazon

Every major AI assistant sees the same click-to-answer shift, but they are betting differently on whether — and how — to put ads inside it. OpenAI turned ads on for US Free and Go tiers on February 9, 2026, a single sponsored card at the bottom of a response; for how that placement is chosen, see the serving mechanism — contextual matching with answer independence — and for the product itself, what ChatGPT ads are.

The most important counter-example is Perplexity. It launched sponsored follow-up questions in November 2024 (at a reported CPM above $50) and then publicly dropped advertising on February 18, 2026, saying it was “in the accuracy business” and moving subscription-only; its 2024 ad revenue was reported at only about $20K. Read alone, that looks like AI advertising failing. Read in context, it is the opposite: the incumbents are all still in. Google has run ads inside AI Overviews since May 2025, Microsoft Advertising has monetized Copilot since April 2025, and Amazon is placing sponsored ads inside its Rufus shopping assistant.

The takeaway is the honest nuance the skeptic needs: Perplexity retreating does not mean the category is collapsing — it means companies with different trust postures and business models are choosing differently on the same shift. The full head-to-head and the incumbent ad-scale numbers live in ChatGPT ads vs Google and social ads.

How Big Could This Get? The Forecasts — and the Outlier

The money is where the story gets misread, so keep the tension explicit. eMarketer forecasts US AI ad spend at $32.03B in 2026, rising to $68.25B by 2030 — but with two caveats that change the picture entirely: 80%+ of that total is “AI-adjacent” (ads placed next to AI-generated content, not inside a chatbot), and standalone-chatbot ad revenue — ChatGPT and its peers combined — is forecast to reach only about $5B by 2030, up from under $1B in 2026.

Against that sits the outlier, and it deserves to be labeled plainly. OpenAI’s own internal projection — reported via Axios and The Information — puts its ad revenue at $2.5B in 2026, $11B in 2027, and $100B by 2030, and a Barclays model reaches $102B by 2030 on the assumption that advertisable users grow from 0.7B to 1.7B and ad revenue per user climbs from $3.50 to $60. That ~$100B-by-2030 number is roughly 45% of Google’s entire current Search ad business (Search & Other ad revenue was $224.53B in FY2025; Meta’s ads were around $195B). It is an aggressive company ambition and analyst model, not a market consensus — and it is never presented as one here.

The honest synthesis: eMarketer’s realistic in-chat read (~$5B by 2030) is smaller than a single quarter of Google Search today, while the OpenAI/Barclays projection is more than twenty times larger. Both are forecasts; the gap between them is the actual state of knowledge. MB Adv Agency reads that gap as a reason to test the channel, not to reallocate a budget on the strength of a 2030 number. The full incumbent-scale comparison belongs with ChatGPT ads vs Google and social ads.

The Trust and Ethics Debate

The part most AI-optimistic write-ups skip is the genuine, dated debate about whether ads belong inside a system people treat as a confidant — and every view here is its author’s stated position, not neutral fact. Zoë Hitzig, an ex-OpenAI economist, resigned and published a New York Times guest essay on February 11, 2026 warning that ads risk “manipulating users in ways we don’t have the tools to understand” and that ChatGPT holds an “unusually sensitive archive.” Writing in Fortune on February 9, 2026, Scott Galloway framed the vulnerability as ads intruding on ChatGPT’s “therapy” use case — for example, a sponsored medication surfacing to a depressed user.

The rivalry sharpened the debate. Anthropic positioned Claude as permanently ad-free in a Super Bowl 2026 campaign, and Sam Altman rebutted it on X on February 4, 2026 — “we would obviously never run ads in the way Anthropic depicts them” — pointing to OpenAI’s stated principle of answer independence: that ads run on separate systems and cannot shape, rank, or alter ChatGPT’s responses. That principle is OpenAI’s stated, unaudited policy, and it should be read as a claim rather than a verified behavior.

Regulators are now applying pressure directly. A 42-state Attorney General probe opened on June 12, 2026 that explicitly names ChatGPT’s advertising practices — alongside engagement, consumer and health data, and minors — as an area of inquiry; it is information-gathering, not a finding. An EU Digital Services Act review and California’s SB 243 add to the backdrop. This is the pillar’s credibility check, and it is why the privacy and regulation crux is covered in full in ChatGPT ads privacy and data.

How Advertisers Should Prepare

The practical close is honest and non-prescriptive. The channel is real and live, but it is an early pilot, so the goal is to learn from it without over-committing to it. Four moves hold up regardless of which forecast turns out right.

1. Test and learn now, but size the bet to the evidence. ChatGPT Ads is live and buyable, so run small, measurable tests — not a budget reallocation placed on a 2030 forecast. Run it the way you would run any test-and-learn optimization in Google Ads, inside a diversified PPC program. 2. Watch the trust signals. Perplexity’s reversal, the 42-state AG probe, and the personalization-and-memory debate all bear on how durable in-chat ads turn out to be, so track them as leading indicators, not noise.

3. Build first-party data. The durable asset across every AI-ad platform is your own customer data and conversion tracking, so invest there first; build measurable conversion tracking with the pixel and Conversions API before you scale spend. 4. Think GEO alongside paid. Being cited in AI answers (GEO and AEO) and advertising in them are complementary, so a resilient plan invests in both — route the organic depth to how AI is changing SEO.

MB Adv Agency treats ChatGPT Ads the way it treats any new channel: a controlled test inside a diversified paid mix, measured against first-party conversions, not a bet placed on a distant forecast. That posture is deliberately boring, and it is the one that survives whichever way the market breaks.

Frequently Asked Questions

Methodology & Sources

This pillar is a balanced synthesis, so it orients and argues and then routes depth to its siblings rather than duplicating it. Every figure traces to a named third party: the market forecasts are eMarketer (via PPC.land reporting eMarketer; granular splits are paywalled) and, for the aggressive outlier, OpenAI’s internal projection reported by Axios and The Information plus a matching Barclays model. The ~$100B-by-2030 number is never presented as consensus — it is a company ambition and analyst model, stated against eMarketer’s realistic chatbot read of about $5B by 2030, and that caveat appears wherever the $68B total or the $100B figure is used. The shift figures are Gartner (traditional search volume down 25% by 2026, a 2024 forecast) and SparkToro/Datos clickstream estimates via Search Engine Land (US zero-click 58.5% in 2024 to 68.01% in early 2026); the zero-click numbers are clickstream estimates, not Google-published. Reported items are labeled as such: Perplexity’s CPM above $50 and its ~$20K 2024 ad revenue. Every opinion carries its author and date — Hitzig (NYT essay, Feb 11, 2026), Galloway (Fortune, Feb 9, 2026), Altman (X post, Feb 4, 2026), and Anthropic (Super Bowl 2026) — and OpenAI’s ad-answer-integrity principles are presented as stated and unaudited. No mbadv client metrics appear in this article: MB Adv Agency has no ChatGPT-Ads service page, and its perspective here is qualitative. Search demand is from Ahrefs Keywords Explorer (US, retrieved 2026-07-01). Reviewed by MB Adv Agency, July 2026.