Switching cost is more than you think. For example, I don't like Zendesk it's expensive and over-engineered. But I stick to it because I know how hard it is to port my docs over and re-setup the live chat. It's half a day work. To save $50/mo, it's not worth my time. My focus is growing my company, not to save $50/mo.
I run SaaS and spent alot in SaaS products.
1) SaaS has a running cost; servers, support, devops, and so on You generally have a minimum cost to run a SaaS business. It's significantly less than it used to be, but it's still there. You have to run servers, unlike iPhone Apps that run on the consumers device.
2) The iPhone and Android App Stores were one of the purest environments for Revenue Management experiments we've ever witnessed.
By way of my background, I worked for a while in the transport industry a few years back, and was responsible for the platforms that sold over 4 million tickets online per year. I worked with revenue management teams, so have some experience in this area. And at the time we would discuss the App Store and pricing trends.
My experience does not grant me authority, these are just my observations and opinions :-)
So back to Revenue Management: The market for smartphones and customers for the App Stores kept growing and the distribution model of the App Stores meant it was easy to run global experiments on pricing.
App Developers quickly realised that their Revenue Management graphs showed increased Revenue and Profit when they lowered the price due to the never-before-seen size of the smartphone market. Suddenly software developers had access to millions of customers around the world who would pay 0.99c for an App.
You could entice new customer with low prices, and at the time it seemed like there was a limitless supply of new customers who would pay 0.99c. So as a business you kept on generating more and more revenue.
You could scale out your business to millions of customers at almost zero additional cost. As long as you didnt care about providing any support of course.
Later everyone realised that Advertisers would pay more, and we ended up with trash Adware Apps.
At a former job, we had a rule that if someone claimed something was easy or trivial or the like, they were assigned to accomplish it.
Best way to do this is a pure commodity play. Like something totally replaceable where the cheapest possible option that works is good. I’ve seen many Indian and Eastern European teams try this. Just google “[Some SaaS] vs” and see the products pop up. It’s possible but it isn’t as easy as it sounds.
Furthermore there are a lot of edge cases and special features you don’t see. Special flags they turn on for customers with some special need, or in some industry. Closing enormous b2b enterprise deals often includes a good amount of custom logic and integration that no one else understands.
Most SaaS is B2B, where software is doing real work of some value, and needs are dramatically more complex, but the addressable market is much smaller. In this environment, contrary to your assertion, adoption and switching costs are significant. While there still may be some network effects, much more dominant is general inertia in business where people just want to get their work done and have little tolerance for change. SaaS takes advantage of these dynamics by handling operations and providing ever more features, capabilities and integrations that cement the software's position inside client companies. At some point it becomes almost impossible to change, even for a software that 80% of the users would prefer, because of some un-handled use cases, or minority of internal stakeholders that would stage a revolt.
Like others have also said, trust plays a key role. Companies usually like to stick with a tool that already works and where they know what they're going to get vs taking a risk on a copy-cat to save a few bucks.
If your market is businesses with employees, the difference between $1/month and $100/month for a piece of software is a rounding error. This means even something that only saves the business an hour each month, is worth it for $100. Even if you buy 100 of these apps, it’s still cheaper than the all-in cost of a single employee.
Conversely, you could ask the same question about anything. For example:
A lot of [employees] are making money hand over fist yet they seem easily replaceable. Furthermore most [employees] don’t have network effects or large switching costs. Given these dynamics, why haven’t we seen a race to rock bottom in [employee] prices?
Either the market is persistently inefficient for some reason (usually isn’t the case) or there’s something you’re not accounting for in your assumptions.
If you’re one of the people who say things like "Intercom? Pffft! I could spin up a clone in a weekend!," I would say you're in the latter.
It is all about the upfront costs and maintenance.
It takes some serious capital and talent to build a high-availablility data center that SaaS run on.
The software is just icing on the cake. You did not include the cake, the plate the cake is on, the table the plate is on, the chair at the table, and the roof overhead. All of those other components are needed and they have costs which must be paid for everything else to work so well and be so easy to use.
Replicate what? The code ? May be. The entire business? Not that easy. Running a saas is lot more then writing code.
Vendor assessment, legal concerns, data privacy concerns, talks about SLA guarantees, talks about 24/7 support plans and much more. There will likely be several departments involved. Technical folks, legal people, data privacy experts etc.
That new deal could pass easily thru 50 peoples desk before getting signed eventually. For what? A 15% saving that could be wiped out with the next round of price adjustments from the new vendor? Simply not worth it. That is why SaaS revenue tends to be so sticky.
It's a similar dynamic to why professional athletes are paid well: they exist in a market where you can't substitute quantity for quality.
In contrast I think the race to the bottom that can be perceived as concerning. Will you still be there, or cause some critical outage?
Can I trust the product I'm using to stick around, to be secure, to not jack up prices?
If the product is clearly a clone of an existing product I'm much less likely to trust it - my very first introduction to them was through an action that doesn't inspire me to trust them!
Most likely your impression is wrong, the easily reolicable ones are probably not making huge profit margins, the ones that are hugley profitable aren't easily replicable, whether for IP or other reasons.
Also businesses recognize, broadly, that when vendors offer something for nothing, the value is commensurate with the price.
Unless you're in a communist country and all.
They aren't, though.
> Furthermore, most SaaS companies don't seem to have network effects or any particularly large switching costs.
On the contrary, switching costs is a big part of what they optimize for.
> Given these dynamics, why haven't we seen a race to rock bottom SaaS prices in and around $1 like we did with games and paid apps?
Games and paid apps that aren't shovelware shit don't cost $1 either (if they do, they're screwing you over with some combination of IAP, ads, and embedded surveillance frameworks).
My take: software resists commoditization. On top of that, SaaS is the ultimate attempt to ensure your software is not a commodity.
Think of it this way: is Netflix a competitor to Disney+? To HBO Max? To YouTube Premium? Only to the extend their catalogues overlap. Which, unsurprisingly, they don't. Since movies/shows aren't good substitutes - e.g. if I want to watch Star Trek, neither Star Wars nor Friends are actual alternatives - this means I have to pay for all of them[0].
It's obvious in case of video streaming, but you can extend this argument to any SaaS. Even for SaaS in ostensibly the same market, they tend to not have full feature overlap, and always find a way to make sure costs of switching are big. No common import/export format. Getting you used to a idiosyncratic workflow, or an interface. Unique integrations. Etc.
Think about a person working in Google's office suite collaborating with a person working in Microsoft Office. It's a painful, uphill battle unless both of them learn both systems and then pick either. But they still need to know both, because they're very much not equivalent and strongly non-overlapping in features.
In similar ways, they're protected from competition. Take Spotify. The streaming music player part is trivial to replicate these days[1] - but this fact doesn't spawn millions of competitors, because music player is not the valuable part of that SaaS. The deals with record labels are, and good luck getting those. Little competition -> not much pressure to lower prices.
--
[0] - Which, in practice, means paying for one or two, and then sailing the high seas for the rest, because ain't anyone has time or patience to deal with half-assed enshittified webshit chrome around a semi-broken in-browser video player. But that's orthogonal to the topic at hand.
[1] - In fact, Spotify has been innovating hard over the last decade to make their player as annoying to use as possible - you really can't make something this anti-ergonomic on your first try.