SaaS to Self-Hosted: Tech Pendulum Shift

There seems to have been a massive shift in the available supply of money for tech companies to burn. The ZIRP era has ended. The next question for me is how this will impact the way we build and sell software.

Enter DHH and Jason

Jason C. and DHH had a very interesting conversation in a recent edition of This Week in Startups regarding the impact of the recent belt-tightening required by the market shift. Many SaaS products tend to charge based on the usage of their product. Pricing is commonly based on how many accounts or the volume of data processed by the application. This cost of the product scales with its usage. Some of these offerings have managed to become deeply integrated into the workflows of their customers, which makes their products sticky. If the SaaS company changes their pricing structure, there is little a customer can do. The customer could find an alternative solution, but that is time and money not spent on its own product and operations.

The solution that DHH proposes is a shift from subscription/usage-based pricing back to the licensing-based model. He is not only suggesting a change in the pricing structure but also a shift from SaaS to self-hosted solutions. This could reduce costs by an order of magnitude. Most differentiated SaaS products cost a lot of money and time to build. This means the price must be proportionally high. Any company that stops building its “moat” will eventually be undercut by a competitor.

AI, Twitter, and Work from Home

The most expensive part of a software company is its employees. Many tech companies pay top dollar for talent in the most expensive city in the world, San Francisco. Once COVID opened the door to work from home, employees and businesses explored other options outside of SF. In theory, companies could hire the same level of talent in another city and pay substantially less for it, especially if the talent is outside of the US. It really challenges the theory of “magic” that happens when talent sits next to each other. I’ve worked from home ever since graduating from college. I have to put effort into learning things from my peers. Fortunately, I was working in an office before graduating, so I was able to pick up a lot of the unspoken rules of working in tech. This type of learning wasn’t something that was often talked about because it was a given that you worked in person with everyone. I am aware that it requires more effort on my part to learn from my peers and have brown bag conversations. Luckily, the people I work with get more interesting the more I talk to them :) All that to say, I think the jury is still out on whether or not there is a difference in hiring talent in a tech hub vs. everywhere else. Like anything else, the answer will be nuanced. The often talked about Musk takeover of Twitter has been used as a case study on how deeply tech companies could cut while maintaining or improving product stability and growth. We’ve seen in the news that many of the major tech companies have cut up to around 30% of staff with little to no issues. Will we see deeper cuts in the years to come? Maybe, but I think we are more likely to see new companies capitalizing on workforce efficiency. Another driver of efficiency will (to some extent already is) is AI-assisted programming. There are many high numbers thrown around about it improving developer output. I’m skeptical of this. I have not yet put a lot of time into AI copilots. I have watched a good number of tech YouTubers using GitHub Copilot. It seems like it’s closer to advanced auto-complete than the singularity.

Tying it all together

DHH ties all of these ideas together by identifying a SaaS company ripe for disruption: Slack. Slack has seen little innovation since being acquired, and there are a litany of competitors doing close to the same thing but going after different segments. On top of this, it’s very expensive. Companies could spend millions just to have text chat with each other. His new company aims to undercut Slack by offering a similar solution at an order of magnitude price difference. They aim to do this by running the company far leaner and building the app to be self-hosted instead of SaaS. This slashes operational costs for the business. This is an incredibly interesting idea, and I will be watching closely to see what happens.