AI era moats may shift from network effects to usage, where data from user interactions creates barriers to entry, enhancing AI product value, especially in vertical applications like Google's search.
Usage is the Moat
- In the internet economy, the network effect was the moat. In an AI economy, usage could be the moat.
- The next trillion-dollar moat may not be built on patents or network effects— it will more likely be constructed from billions of user interactions that your competitors can't replicate.
- A moat in business is generally referred to as some significant barrier of entry that prevents competition from coming in and eating your lunch.
- Each new technological revolution has new potential moats in them. In the advent of telephony, we saw a very powerful moat - the network effect.
- This AI era poses a new question: can usage be a moat. In particular, can the data collected by customers using a product create a sustainable moat.
- Perhaps the biggest moat of this kind in history is from Google. Google initially was able to automate search with their PageRank algorithm. However, the real advantage that they had was all of the click data that they were gathering from previous searches.
- The intent behind ChatGPT was less to create a gangbuster consumer product and more to get this kind of usage data feedback.
- More data can mean a better AI product.
- I assert that the value of a product increases logistically with the amount of usage on the platform.
- The moat is in creating a feedback loop from individual customer-specific usage and their specific problem. One could argue that this effect is greater in vertical applications where this data is the most specific and proprietary.
- In this era that seems to be led by artificial intelligence, moats might be different. We have to ask the question of if a moat has changed in this period of time. Perhaps, usage is the moat