Has AI removed the appeal of vertical SaaS?

Let's put aside for a moment the question of 'is AI bad for humanity' and accept that it is a thing that exists in the world. Particularly in the world of 'tool that writes code'. Right now, for circa $200 a month you can buy a tool that can write you some code. It wont be perfect, but if what you're doing has been done before , you'll probably be fine.

The reality for almost all the vertical B2B apps in the market today is that most of what they do has been done before. What was once a moat, building a very specific product for a narrow ICP, may have started to dry out. Is Vertical SaaS dead? Maybe not, but it's not as healthy as it used to be.

Talking this week with some tech VCs, there's less and less desire to invest in Vertical SaaS that doesn't have some other form of unique IP. The risk of competition is too high.

Those feeling this the most are the mature vertical SaaS companies which may struggle to raise their next round and may not have the runway to develop something unique. We probably wont see the fallout of this for 12 to 18 months but the early signs are there.

Once upon a time (early last year), if you knew plumbing and how to code you could build a SaaS for plumbers and be confident that you'd do okay. You'd embed plumbing specific workflows and sell to plumbers because you spoke their language. You could rest easy knowing that there are plenty of trades and the chances that you'd be flooded with competitors in 'plumbing' was low. If you became the go-to product you'd get good market share and build a healthy business.

As of today, some of those assumptions are no longer valid. Unless you have a reason beyond 'we customised things to a niche' you may not have a moat any more. Anyone with a tiny AI budget can likely spin up a generic competitor to your plumbing SaaS pretty quickly. The current set of AI tools is very good at 'do things that are done a lot.' Most B2B SaaS code falls directly into the bucket of 'things that are done a lot'.

This isn't uniquely an AI thing. If we had ten times as many startups the competition would probably naturally arise anyway. What's new is that becoming a viable competitor has become much, much faster.

Is vertical SaaS doomed? Probably not in general, there's a need to find a moat that is deeper than 'nobody else will put the effort in to doing this.' The market is about to be flooded with products that get most of the way there. If that was all you had to go on then yeah, you're probably in a tough spot. On the other hand, if you have a real moat you'll likely do better than ever.

AI has made the bottom 80% more accessible than it's ever been. The top 20% is still there for the taking. If you're slogging through the work to get to the 80% mark or you have a large existing code base serving that 80% functionality, you might be falling behind. The top 20% however, will still be won by people willing to put in the work.

If you spend a lot of time with today's AI tools you'll find they start to fall over when they get too far from the median. All the work that's pushing the envelope, the deep-tech, the regulatory compliance, it's all still there. Investing time in those things will still help you build a winning product. Spending that time instead on building the 'easy' 80% when your competitors are automating the boring bits is a trade off that makes less and less sense.

This is not at all a post suggesting we no longer need developers. The delta between AI code created by devs and non-devs is big. All those 'one prompt, one product' tools are great for a demo but you'll still need devs. Those devs seeking the fastest path to an MVP that isn't built on sand can get a product in front of users significantly faster than ever before. Even if you end up with a rats nest of code and never make it past the 80% mark, having more competition at that level is going to drive down what consumers are willing to pay. Features become become a commodity.

The drop in 'effort' needed to get an okay product into the market also changes the weight of factors like marketing, distribution. All the non-code reasons that people buy a product. The battle for market share will be won with a combination of one or two key technical insights plus a whole lot of sales and marketing effort. This is the way that the battles for market share have been won in less niche products for a long, long time.

None of this is to say that if you know plumbing and how to code that you should give up on building a Vertical SaaS. You're just going to face a lot more competition than before. You need to find your unique, hard to copy, edge to make sure you're not a few prompts away from being replicated.