In the AI era, you don't click on ads. You talk to them
OpenAI's Sponsored Agents try to stretch the unit of advertising from impressions and clicks to conversation and transaction potential. It won't topple Google search ads overnight, but it could change how advertisers capture purchase intent in the first place.
Google's real asset was never the search box. It was the moment it caught intent
For 25 years, the economics of search advertising stayed simple. A person types "hotels in Seoul," "car insurance," or "best laptop." Google guesses what that person wants right at that instant. Advertisers pay to show up in front of that intent.
Search ads are powerful not because Google owns a lot of web pages, but because they catch the exact moment a person is more likely to spend money. Google's Search & Other ad revenue came in at $50.7 billion in the first quarter of 2025, $54.2 billion in the second quarter, $56.6 billion in the third quarter, and $63.1 billion in the fourth quarter. Add those up and you get roughly $224.6 billion for the year.
What's more, fourth-quarter 2025 Search & Other revenue grew 17% year over year. That's a hard number to square with any claim that AI has already weakened Google's search-ad business.
What's different about ChatGPT is that its "intent sentence" is far longer than a search query
Say a user writes something like this:
"A couple in their 50s is staying five nights in New York in November. Find a hotel that's not too expensive, close to Central Park, and quiet."
Information that Google Search would typically split across multiple queries and filters sits in a single sentence. Travel dates, city, length of stay, price sensitivity, preferred neighborhood, even a preference for quiet. In ad-industry terms, that's extremely dense commercial intent.
That is the strategic significance of Sponsored Agents. OpenAI now lets users, after seeing a relevant ad, open a separate conversation with an agent that is clearly labeled as sponsored by a company. Users can ask follow-up questions about a product or service and, when they're ready, click through to the company's website.
Impression → Click → Website (the last 25 years)
Conversation → Qualification → Transaction (the next step AI could open up)
The key point is that the last step, "transaction," isn't a fully built feature yet. Right now the flow ends with a conversation followed by a link to the company's site. But if product data, inventory, CRM and payment systems get more deeply wired into these agents, advertising could shift from simply buying traffic to something closer to "a consultation with an AI sales rep."
An AI that shows ads and an AI that makes ads have arrived at the same time
Alongside Sponsored Agents, OpenAI has also expanded its tools for advertisers. Through Ads Manager inside ChatGPT Work, advertisers can build and edit campaigns in natural language and analyze performance. They can turn a website or a campaign brief into an ad and get suggested copy and images.
HubSpot is the first CRM partner, connecting ad creation, lead capture, follow-up workflows and performance analytics inside the CRM itself. Shopify is the first e-commerce partner. US Shopify merchants can link their product catalogs and measurement tools to ChatGPT Ads, with international rollout set to expand to more ChatGPT Ads-supported countries starting September 23.
This matters because the ad platform isn't just changing the interface users see. AI is rebuilding the advertiser's own workflow, from buying and creating ads to targeting and managing leads.
That doesn't mean Google gets pushed aside anytime soon
| Metric | Value |
|---|---|
| 2025 Google Search & Other ad revenue (simple sum) | $224.6B |
| 2025 Meta ad revenue | $196.2B |
| 2025 Amazon advertising services revenue | $68.6B |
OpenAI said that by August 31, 2026, before ChatGPT Ads even hits 200 days since launch, it had already crossed $1 billion in annualized revenue. The growth rate is fast, but the absolute scale still isn't close to Google, Meta or Amazon.
The more important question isn't how quickly ChatGPT ads grow, but whether advertisers can reproduce the ROI they get from Google Search inside ChatGPT. Search advertising spent decades building auctions, quality scores, conversion measurement, advertiser tools and payment infrastructure. Conversational AI advertising is still at the early stage of building that same system.
The four companies hold different kinds of "intent data"
| Platform | Core data | Strength in advertising | The AI-era question |
|---|---|---|---|
| Search intent | Explicit queries right before a purchase, overwhelming ad infrastructure | Can AI Mode and Gemini keep holding commercial search intent inside Google? | |
| Amazon | Actual purchase history | Very short distance between product search and checkout | How much can a shopping agent automate, from recommendation to order? |
| Meta | Interest, relationships, content reactions | Ability to discover and manufacture demand | Can AI convert a user's interest into a commercial conversation? |
| ChatGPT | Context and conditions inside a conversation | Richer context and follow-up questions than a search query | Can this context be turned into ad value without damaging trust? |
For investors, unit economics matter more than the launch headline
Alphabet investors are probably better served watching the growth rate of commercial queries and advertiser ROI in Google Search than the headline numbers coming out of ChatGPT Ads. Google itself says AI Overviews and AI Mode, now integrated into search, are generating new commercial queries. The fact that Search revenue is still accelerating is strong evidence against the "AI is already eating search ads" narrative.
For Meta investors, what matters is how much AI lowers the cost of creating ads and targeting them. Meta's 2025 ad revenue came to $196.175 billion, up 22% year over year. If OpenAI's new model proves itself, Meta has a growing incentive to bring conversational advertising into its own messaging apps and AI agents.
Amazon looks like the most natural player to both defend and attack here. Its 2025 advertising services revenue was $68.635 billion. Amazon already holds purchase history, product catalogs, pricing, inventory and payments inside one system. As AI agents evolve toward merging advertising with transactions, the value of that data combination could keep rising.
Insight Times Editorial Desk
