Monetization in the Consumer A.I. Era

Advertisers really only pay for two things, attention and intent. Attention is often associated with longer term seeding of a brand message, hopefully recalled by a consumer when they are in the market. Intent is the opposite - people that seem to be in the market right now. Whereas brand advertising hasn’t changed much in 50+ years, performance or direct response advertising as we know it has only been around a couple of decades, fueled by connectivity, devices and data. This has enabled businesses to find customers for just about anything, very quickly. There is an always-on debate regarding the efficacy of brand and direct response advertising (see the recent 耐克 hysteria) but regardless, nearly all new businesses are compelled to generate sales as fast as possible. But, it is also in their interest to stand for something and begin crafting (and potentially spending on) a longer term brand position.?

Generative AI tools are now emerging, offering capabilities such as creating text, images, and videos, as well as answering questions and responding to search queries. As these consumer products proliferate, the question of monetization arises. Many of the tools aim to find signal in the vast noise of the Internet, providing users with the most relevant information. By design, this should mean less scrolling, fewer clicks on links or images, and more efficient information retrieval and presentation. The more efficient at understanding, generating or taking action, the less time people will spend.?

For all but pure entertainment services this seems certain to mean less of an opportunity for brand advertisers to share their messaging - good timing perhaps for the rise of streaming TV (“streaming accounted for 40.3% of total TV usage, the highest share ever recorded”, 尼尔森 , June ‘24). What about the opportunity for direct response??

Many here seem eager to write off 谷歌 (e.g. costs more to generate results than show links, won’t want to cannibalize their existing revenue streams, etc.) but they have vast infrastructure advantages. These range from user data to ads products to measurement tools and of course more recently, enormous resources dedicated to A.I. You can see from the image below that while generative search results are a new experience, the leap to monetization at scale isn’t that difficult when all the other pieces are in place. In this example I believe the products are displayed for free as the businesses are sharing their catalog with Google for shopping purposes. That said, you can imagine this being a paid opportunity before too long.?

To compete, start-ups need not only build a compelling product and attract users, but also put together the elements that enable them to understand and execute on intent. Without it, monetization will be a poor user experience in addition to being extremely difficult.?

Outside these traditional paths perhaps there are others not yet conceived. For example:?

1. Agent marketplace: When completing tasks for people, service agents will have to interact with other premium services. This itself could work as a marketplace with revenue share, the spoils going to the service that aggregates the most demand.?
2. Vertical integration: Taking this a step further and for the most valuable transaction categories, some of the more well funded may go beyond calling and scraping and generating. Google did something like this in 2011 when it bought ITA for $700m. Lacking visibility into the global airline timetables and pricing, Google gained the infrastructure enabling the consumer facing flight search and price visibility experience to improve dramatically.