Last week, Reddit filed their S-1 to go public. At least 10% of their revenue - about $60m - comes from selling data to train Large Language Models. Reddit’s data sales revenue will likely be much more than 10% by the end of the year. Quoting directly :

We expect our growing data advantage and intellectual property to continue to be a key element in the training of future LLMs.

This raises a fundamental question : What if the revenue from data sales dwarfs the revenue from ads?

LLMs need data. They compress this it & reconstitute it to answer user queries. At Reddit, like many sites on the Internet, the content changes often. Users want to search the data for product reviews, travel recommendations, facts, & fun (the latest memes). Google isn’t the only one after this data : OpenAI & other providers are also brokering direct deals with Internet publishers.

Because the data needs to be fresh, Google & others will continue to pay for access & potentially increasingly more for exclusivity, low-latency, or data of a particular kind that improves the accuracy of their models.

Data sales invert the business model of the Internet.

Instead of Reddit building product experiences that create good advertising data to earn more on ads, Reddit will launch product experiences that produce more valuable data to feed to LLMs. The LLM vendors should pay more for better data.

User data remains the currency of the realm, but it’s packaged & sold in a very different way. Cookies, the moribund technology that created the ad world, will be replaced by data-purchasing contracts.

One day, we may visit websites that have fewer ads or none at all. The revenue model of the Internet will have changed. Publishers’ sell the data directly to the search companies.

Should we evolve this way, there are any number of questions around user privacy, data controls, regulation, not to mention how product experiences themselves may change.

The days of the ad network may be numbered & with it, an era of a new business model for the Internet.