AI Search Ads Face a Difficult Road to Scale

概览

This episode of Marketplace Tech examines why advertising, the business model behind much of the internet, may be harder to translate into AI search and chatbot platforms than it was for search engines, social media, e-commerce, or mobile games.

The central discussion is with Garrett Johnson, a marketing professor at Boston University. He says AI companies face competition for users, competition for advertisers against mature platforms like Google, Meta, and Amazon, and the operational challenge of building ad systems that can scale.

A major conclusion is that AI platforms may delay ads while competing for users, but advertising is likely difficult to avoid because it pays for expensive services and can help users with shopping-related tasks. The episode also connects AI ads to the shift from search engine optimization to generative engine optimization, where being mentioned by chatbots may become as important as ranking in search results.

分段落总结

[00:00] Sponsor Message: Tomorrow’s Cure

[事实] The episode opens with a sponsor message for Tomorrow’s Cure, a Mayo Clinic podcast about technology and medicine.

[事实] The ad says the podcast covers AI-powered diagnostics, cancer therapies, surgical technologies, and a season premiere about carbon ion therapy.

[推测] The sponsor placement frames the episode for an audience already interested in technology’s impact on major industries.

[01:05] Why AI Search Ads Are Not Straightforward

[事实] The host introduces the question: if the internet runs on ads, why not AI?

[事实] The episode notes that Perplexity reportedly pulled back from plans to integrate ads into its AI search engine.

[事实] Internal documents showed Perplexity made only $20,000 in ad revenue in the fourth quarter of the previous year.

[推测] The example is used to show that adding ads to AI search is not simply a matter of copying older internet business models.

[01:55] Three Main Challenges for AI Advertising

[事实] Garrett Johnson says AI ad platforms face intense competition for users, possibly more than earlier search and social platforms did.

[事实] He says AI platforms also compete for advertisers against mature digital ad businesses such as Meta, Google, and Amazon.

[事实] He identifies scale as another challenge: platforms must onboard many advertisers, including small and medium-sized businesses, and build a personalization flywheel.

[推测] Partnerships such as OpenAI’s with Walmart and Shopify may be valuable because they can provide conversion data about what consumers like.

[03:00] AI Platforms Must Start Simple

[事实] Johnson says AI companies may need to begin with familiar ad formats before evolving toward broader or more AI-native ad offerings.

[事实] He says platforms need to meet users and advertisers where they already are.

[推测] Early AI advertising may resemble existing search or social ads before companies discover formats better suited to conversational interfaces.

[03:22] User Growth Versus Ad Revenue

[事实] The host asks how competition among OpenAI, Google Gemini, and other AI platforms affects advertising rollout.

[事实] Johnson says companies face a strategic dilemma between investing in user growth now and investing in ads that could generate revenue today.

[事实] He says companies that want to do advertising probably need to start sooner rather than later because of returns to scale.

[推测] Delaying ad development could leave AI companies further behind Google, which already has a major advertising infrastructure advantage.

[04:42] Ads May Be Inevitable

[事实] Johnson is skeptical of claims that companies will permanently avoid ads.

[事实] He says companies from Netflix to OpenAI may dislike or downplay ads at first but eventually use them because ads pay the bills.

[事实] He also says users of general-purpose platforms often want help buying things, and advertisers have knowledge about meeting consumer needs.

[推测] The interview suggests advertising could become part of AI platforms even if companies initially frame it as a lower priority.

[05:41] Will Users Reject Ads in Chatbots?

[事实] The host notes that many consumers may currently enjoy chatbot search and recommendations because they do not include advertising.

[事实] Johnson says users are primarily using chatbots because they provide powerful, distilled answers to questions.

[事实] He argues that ads are often less intrusive when they are text ads or product listing ads, and that ads can sometimes be useful.

[推测] The risk of being the first AI platform to introduce ads may be real, but Johnson suggests the business pressure to monetize non-paying users is also strong.

[07:06] From SEO to Generative Engine Optimization

[事实] The episode discusses how traditional search created the search engine optimization industry.

[事实] Johnson says businesses are shifting from trying to rank at the top of search results to trying to be mentioned in AI chat responses to related queries.

[事实] He says the principles of SEO and generative engine optimization are similar: create quality content for user queries and have the brand discussed by high-authority websites.

[推测] Generative engine optimization may not replace SEO entirely, but it changes the goal from visibility on a results page to inclusion in a chatbot answer.

[08:15] Fewer Results Could Strengthen Winners

[事实] Johnson says chat results may list fewer options than traditional search results.

[事实] He says this could strengthen winner-take-most dynamics.

[事实] He gives an example where a generative engine might show four pizza options, with three organic listings and one sponsored listing.

[推测] If AI interfaces offer fewer visible slots, smaller brands may have stronger incentives to pay for placement while dominant brands may appear organically.

[08:57] Ad Design Will Shape the Market

[事实] Johnson says the exact design of AI ad placements will have enormous consequences for the ad market.

[事实] He says it is still unclear how these AI advertising systems will look.

[推测] The uncertainty around format, labeling, ranking, and sponsored placement means the economics of AI advertising are still unsettled.

[09:20] Episode Credits and Post-Roll Promotion

[事实] The episode identifies Garrett Johnson as a professor at Boston University.

[事实] Daniel Shin produced the episode, and Megan McCarty Carino hosts Marketplace Tech.

[事实] The transcript ends with a promotion for How We Survive, a podcast about climate solutions and geoengineering.

播客点评/总结

This episode is valuable because it treats AI advertising as a business-model and market-design problem, not just a technology feature. The strongest part is Johnson’s explanation of why scale, advertiser onboarding, user competition, and conversion data all matter.

The discussion is concise and practical. It connects familiar internet advertising history to newer AI search behavior without overstating what is known. The comparison between SEO and generative engine optimization is especially useful for businesses thinking about visibility in chatbot answers.

[推测] The main limitation is that the episode does not deeply examine user trust, disclosure rules, or how sponsored answers should be labeled in conversational AI. It also does not include perspectives from AI companies or advertisers directly.

[推测] This episode is best suited for listeners interested in AI business models, digital advertising, search, marketing strategy, and the commercial future of chatbot platforms.