SaaS profit is market inefficiency, and it will end

I don't think the profitability of "software as a service" businesses is going to last, in the long term (on a timescale looking at, say, 100 years from now). I wrote before about machine-owned enterprises competing the profit out of SaaS, but even before that happens, humans ought to be able to compete most of the profit out of it.

Why it's a market inefficiency

There are almost no barriers to entry in starting your average SaaS business. There are no licensing or other onerous regulatory requirements which might help sustain a monopoly, the startup and operating costs are both extremely low, even negligible (just a VPS and a domain name will do), and the marginal cost of providing service to an additional user is practically nothing. Despite all this, margins tend to be huge, implying an excess of demand or a shortage of supply.

If we assume the only barrier to entry is that you need to be able to write the software, then the only thing standing between any arbitrary SaaS company X and complete irrelevance is that nobody else has yet implemented a cheaper competitor Y that is good enough to convince X's customers to switch.

In the short term there is more to it, of course. If I'm already a customer of X, I'm probably not going to switch to new upstart Y just to save a small amount of money. Not until the new upstart has been around for long enough to prove they're not going to disappear. There is also the problem of marketing: I need some way to learn that Y exists, and to be confident that Y is actually going to solve the problem at least as well as X.

But in the long term that doesn't matter. In the long term, all existing customers of all existing SaaS services will be dead, so their biases are irrelevant, and the good businesses ought to spread organically.

I contend that, in the long term, the businesses offering adequate service at the lowest price will "win". I also contend that, due to the lack of barriers to entry, and the (presumed!) increasing technical literacy in the world, there will be lots of people eager to eat the lunches of everybody over-charging for basic SaaS products, and succeeding, and I therefore contend that eventually prices must come down until it's no longer worth competing.

Essentially, SaaS straddles the gap between a post-scarcity supply chain and a rest of the world that is still used to paying lots of money for things. But when you're selling "goods" that you can get for free, and that anyone else can get for free as well, you can't charge very much without driving customers to your competitors.

None of this is to say that you shouldn't start a SaaS project. You should! If you spot a market inefficiency you should take advantage of it as quickly as possible, before somebody else does.

What can you do to remain competitive?

Apart from the obvious (drop your prices when a credible competitor appears), you can make yourself a less appealing target for competitors by doing things that don't scale, and making those things a core part of your value proposition.

The other option is to just not worry about it and hope that everything goes well for long enough that you'll be dead before it matters. This is my personal preference.

Why I could be wrong

If the space of all possible SaaS businesses remains roughly where it is today (obviously not the case), I think the arguments above hold up: prices must come down as markets get more crowded.

But this is arguably the most interesting part:

New inventions are made all the time. If we assume that the space of known inventions is (roughly) a hypersphere in some many-dimensional space, and that new inventions build on known inventions to expand the hypersphere slightly in some direction, then the set of all possible new inventions is the "wavefront" of this ever-expanding hyperglobule. (I first read this concept somewhere else, which I'd like to link to, but unfortunately I can't find it right now).

In the event that the wavefront of all possible new inventions expands at a faster rate than the software implementation capability of the world grows, then there will always be new possibilities available, and new market entrants will be better served by entering a new niche than by trying to compete on price in an existing niche. That would be an excellent outcome.

I wish I knew what was going to happen.