/ai-search739 words

AI-Generated Chatbot Ads Draw Double the Clicks, Koah Data Shows

AI-written chatbot ads draw double the clicks of human-made creative, Koah Labs says, as OpenAI tops 100,000 ChatGPT advertisers and Google expands AI Mode ads to 10 more countries.

Ads in Chatbots Have Arrived. Here’s What’s Working—and What’s Not
Ads in Chatbots Have Arrived. Here’s What’s Working—and What’s Notstevepb / pixabay
  • AI-composed chatbot ads generate double the clicks of human-made ads, Koah Labs CEO Nic Baird says
  • OpenAI has more than 100,000 advertisers in ChatGPT since ads launched in February
  • Google expanded AI Mode ads to 10 more countries beyond the US
  • Koah reports ~3% click-through on chatbot ads, on par with Google's mature ads system
  • Koah estimates 1 in 4 brand-related chatbot queries concerns returns or support issues

Ads partly written by AI generate double the clicks of fully human-made creative inside chatbot apps, according to Nic Baird, CEO of San Francisco ad-tech startup Koah Labs — one of the first concrete performance signals from the emerging market for advertising inside AI conversational products.

The numbers behind the market are already substantial. OpenAI has signed more than 100,000 advertisers in ChatGPT since ads debuted there in February. Last week, Google expanded ads in AI Mode — its search chatbot — to 10 more countries beyond the US and said users "find them highly relevant." Advertisers remain unsatisfied with the experience but expect it to improve.

Anthropic is the holdout. The company reiterated its Super Bowl pledge to keep Claude ad-free, and its recent IPO filing didn't mention selling ads against the chatbot. That stance is easier to hold with roughly $112 billion in investor cash. For smaller platforms without venture largesse or tech-giant revenues, ads may be the only way to defray the high cost of generating AI responses.

What's working in chatbot ads?

Koah brings ads to more than 200 apps, including DeepAI, a creative-queries tool, and Liner, which adds citations to AI answers. It works with brands such as telehealth giant Hims and flower delivery service UrbanStems. Its advertiser base is about a tenth of OpenAI's, but Baird says the early lessons are clear.

Machine-crafted copy wins. AI-composed ads pull double the clicks, likely because generated marketing text feels more natural beside a chatbot's output. Because Koah's ads are grounded in advertiser-supplied product details and imagery, users aren't hitting hallucinated ad claims. Regulated advertisers complicate the picture: drugmakers, bound by government rules on what ads must or must not say, keep more human control — and get weaker results, Baird says.

How does targeting differ from search?

A Google search for "help with stress" reliably surfaces therapist ads. Chatbots allow looser matches: spa ads perform well in Koah's network after prompts about "stress." Koah selects ads based on "not only the prompt, or the Google search, but also the response from the AI platform, which brings context from the rest of the conversation thus far," Baird said. Chatbot prompts run at least double the length of Google queries, giving the system more signal to work with.

The engagement numbers hold up against mature search. Koah says users click on about 3 percent of its ads — roughly on par with Google's established ads system.

What happens to conversation data?

Koah generally receives a user's entire conversation up to the point an ad is shown. People treat chatbots as confidants as well as assistants, which raises privacy stakes as personalized ads proliferate. Baird says Koah anonymizes queries and stores them in a de-identified way, but declined to detail collection and storage practices. Conversation history could later power personalized post-click landing pages and retargeting — repeating ads to the same user — though Koah has no data on whether people actually want those customized pages.

Where's the risk?

Koah estimates one in four chatbot queries about a retailer or brand concerns a return, exchange, or other support issue. Baird expects manufacturers and retailers to eventually target those users with warranty reminders and AI-intermediated support, steering them away from costlier customer-service calls. Best case: faster resolution. Worst case: a wasted ad buy and a more irate consumer.

Agent-driven commerce looms over all of it. None of Koah's current apps offer autonomous shopping agents, but apps like Instinct and Meta's Muse are starting to take over shopping chores. Meta CEO Mark Zuckerberg said Muse could become the company's biggest revenue source — via transaction fees, not ads. Early studies suggest ads can still influence AI agents. Baird expects to give "the user's agent the best possible recommendation" through discounts and relevant products, provided sponsored content is made "abundantly clear" or platforms risk losing user trust.

Skepticism persists inside the network itself. Kevin Baragona, CEO of DeepAI, says ad revenue has been meaningful but he's not committed to it because the experience doesn't feel right on privacy and design. "I don't think the ads are working terribly well for anyone," he said.

Watch whether more AI developers follow Anthropic's ad-free stance, and whether click-through rates and privacy norms hold as conversation-based targeting scales.

via wired.com (Source)

More from James Calloway

James Calloway

Show full bio

Market editor covering money, funding and market shifts around Search rankings and algorithm updates, generative engine optimization at SERP Journal.

220 articles