In the Age of AI, Build "Platforms," Not "Products"
Join the discussion on Twitter
With AI becoming the latest “THING” in the tech sector, some are predicting an extinction level event for existing software companies. As Dave Yuan of Tidemark points out, this is only partially true. The latest AI tools not only have the power to make software more powerful, but also to greatly decrease its cost of production. With that, many are concerned that legacy platforms unable to leverage AI will be rendered obsolete, and they’re not wrong! If I were a generic workflow product, I would be terrified. The moat around these products is now next to nil as the barriers to entry of software production have been washed away. What were once successful niche software product companies are now being obliterated by new entrants leveraging the functionality of AI models. Perhaps even more daunting is the speed and efficacy of this destruction, with companies that have existed for years and taken millions of dollars to build being dismantled by teams capable of using AI to launch comparable (and arguably better) products in weeks with incredibly lean engineering teams.
Simply put, a lot of legacy software product companies will die and I imagine that a lot of these new entrants will as well, as later entrants come chasing them soon after.
As the barriers to software production come down and competition increasingly comes to market, the inevitable state is a crash in the price of software products. We’ve seen this evolve over the last several decades (in accordance with Moore’s Law) as functionality that once garnered premium pricing is either rendered obsolete or given away at progressively lower prices. After all, can you imagine what folks would have paid for the functionality of Dropbox, Airtable, or Figma two decades ago when they were happy to shell out a couple hundred dollars for basic word processing software?! This is incredible for consumers and customers. We will see amazing new levels of functionality at low prices in the coming years. While benefiting consumers and customers, this will likely make building software products progressively less lucrative over the long-run (competition is bad for business), creating some turbulent conditions for emerging software founders. As we wade through these waters, we’re reminding founders that it’s more important than ever to build PLATFORMS, not PRODUCTS.
The term “Platform” is thrown around a LOT in startup circles, but for the purposes of our discussion, we’ll oversimplify to define it as a software system that generates increasing returns to scale via proprietary Data, Distribution or Development (the 3Ds):
Data — Companies achieving data network effects via 1st party and/or UGC, enabling services with monopolistic information rights
Distribution — Companies achieving customer network effects via the networks of their stakeholders, enabling monopolistic discovery and transaction efficiencies
Development — Companies attracting 3rd party developers to create and market services within the platform, enabling services with monopolistic capabilities and cost structures
Internally, we refer to these mechanisms for increasing returns to scale as “Capabilities” — they are amongst the most powerful forces for value creation and moats. Platforms leverage the flywheels of these 3Ds to grow more powerful with each interaction they have with their stakeholders, ultimately establishing and expanding their moats over time. Historically, this has made platform businesses far more valuable than product companies — they can offer better information without purchasing third party data, they can attract customers more cost effectively and they can even improve the speed or cost of their product development — all generating significant advantages in cost, performance, and profitability over competitors.
While platform businesses were always more valuable than product companies, the mindset must now be equally offensive AND defensive given the progression of AI. Platform companies don’t win on product alone, they win on the networks they establish within those products (i.e., data, distribution, and/or development). While AI is obliterating the value of the technical products, it’s yet to touch these networks making them the only defensible component of a software company. Simply put, it’s the networks that are creating and maintaining value, NOT the actual products which are increasingly at risk of commoditization.
While this can happen in horizontal applications (Slack, Figma and Salesforce are great examples), we believe the opportunity is particularly exciting within the verticals we invest in at Equal (Insurance, Supply Chain, Commerce Enablement and Energy). These industries are still in their early stages of digital transformation and while products exist in these categories, networked products (i.e., platforms) generally do not. This dynamic creates incredible opportunities for our companies to establish themselves as THE platforms in their respective markets, and we’re seeing many of these companies now leveraging AI to accelerate their capture of these industries. Where product companies are seeing an extinction level event, platform companies are now seeing a massive expansion in the productivity of their proprietary data and an equally massive acceleration in their ability to ship products. This in turn is enabling platform companies to build their customer and partner networks even faster than they were able to before. In some cases, the process to forge the networks within their products has taken years, leaving them with little competition in the market today and limited threat of new entrants given that the winning networks that have already been formed. This is a theme we are tracking closely across our sectors as we hunt for additional opportunities and we suspect that it is highly applicable in other legacy markets as well.
Regardless of whether you feel AI is exciting or terrifying, it’s coming. I, for one, couldn’t be more excited for this impact on the companies and industries in which we invest. These categories represent tens of millions of workers who have been left outside of the digital age. This is our opportunity to change that and accelerate their digital transformation. That has the potential to be incredibly impactful — given the size of these markets (each of which is $1T TAM in the US alone), the economic opportunity behind AI-driven platform expansion in these markets may very well be one of the largest prizes in technology today.
Platforms are hard to build, but given what’s at stake, swinging big is the only at-bat worth taking. In the age of AI, build platforms, not just products.