When Angela and I were pitching our newest fund to investors, we started using the term “smart SaaS” to describe the kind of SaaS companies we like to invest in.

Before the SaaS explosion, software and data were only “on premise.” Then, SaaS came along and moved everything to the cloud, boosting efficiency and collaboration since everything was now digitalized and easily accessible. For us “smart SaaS” represents the next stage in this evolution. Now that data is on the cloud, products can begin to leverage it and use AI/machine learning to create a 10x better user experience.

Smart SaaS products can automate processes and take care of all the things that we don’t want to do, while also having the potential to be intelligent in the future and create interesting data network effects along the way.   

When we consider SaaS products in virtually every vertical, we realize how much potential there still is to re-invent the user experience. For example…

• Why can’t accounting products do a better job of reconciling bank accounts?
• Why aren’t there better ways of automating standard customer interactions? (Note, this is something our portfolio company Ada is actually starting to address very successfully)
• Why can’t HR systems understand who in the organization has what knowledge and automatically build a skill/talent graph?
• And across the board in every vertical, why is a/b testing still so manual?

The biggest challenge for any start-up trying to fix these problems is that most of the data needed to build a smart SaaS product is stored in legacy systems that have no intention of making this data accessible to a future competitor.

Therefore, entrepreneurs need to figure out how to bootstrap the chicken and egg problem and overcome this lack of data. There are a couple of ways to do this:

• Solve a specific problem so well from the beginning that it doesn’t matter if you don’t have much data for the first few years,
• Make humans part of the workflow and have them feed data into the product. Over time, the “machine” will get smarter and smarter and you can rely less and less on human intervention.
• Try the AI on publicly available data and then augment it with private data as customers start using the product.
• Build a product on top of more “open” legacy systems that have APIs. The key question here is whether you can become a standalone company when you’re just a layer integrated into another system.

This is an exciting time for SaaS startups. There are countless opportunities for the next generation of enterprise apps to us AI/machine learning to become smart – “smart SaaS” as we like to call it.