The Complete Guide to SaaS Business Models

With SaaS proving itself to be here to stay, and the novelty having given way to pure development, adoption and innovation, one would expect to see all manner of new and clever SaaS business models come up as competition and the struggle for profitability and ease of customer acquisition increase exponentially.

But, since the last time we talked about this topic, there have been no real changes nor innovations regarding SaaS business models. So, this tedious topic is once more going to be tedious, as we yet again go over the main models used in current times. We’ll spread this out a little more than we usually do, because there’s such an absence of new things to say, that maybe we can get a few things out we’ve just not had time to in the past.

How We’re Dividing:

We’re dividing out SaaS business models not just based on intrinsics like we have in the past, but also upon how the subscriptions work where they are present, which means we are going to see five models in use, though three of them are variations on a theme.

We’ll look at each one, cite its strengths and weaknesses, and then you can be the judge on whether or not it’s suitable for your needs. There is no particular meaning to the order these are being looked at in, by the way.

The Ad Revenue Model:

This is the simplest one, and also the most difficult to make profitable even on paper. A lot of startup SaaS concepts have made their initial profit off of ad support, and a lot of hobby projects have seen furtherance over this concept as well.

The problem with it is that first of all, ads are annoying to users, even when they know the reason they are unavoidably there. It’s also a problem that ads aren’t worth much per viewing, meaning clicks are needed if you want it to really take off, or have a huge user base.

A variation on this not worth its own point is the donationware concept, which just appeals to the altruism of all users to donate what they can if they find the service useful.

A lot of SaaS uses this model, but it’s the fringe or general purpose sort of stuff, rarely heavy business computing solutions. Another enduring problem with this is finding the right balance to not inspire your users to block your ads, as well as choosing the right kind to minimize the level of annoyance you expose your customers to.


This is a step up from the ad supported model in that the base functionality is still free. The difference here is that there is an upsell aspect, to get users to subscribe (for money) to a better tier of service, or more functionality than is already available.

The problem here is overhead, as the paying customer support the free users unless the free users are picking up their own slack somehow. Peppering in the ad model on a smaller scale here might solve that problem, and buying out of ads to become a paid user might just be the best motivation for an upsell ever. Or, it may just alienate everyone, there’s no being sure which will happen.

Freemium is hard, and it requires you to be obnoxious, because you have to work to upsell to people, which is always a nuisance to the customers.

Pay Per Use:

Pay per use grants you one time or short time access to computing solutions for an up front payment, not unlike pay per view services and the like over media. If it’s cheap, and your service is worth it, this is actually a very good model to use, but it’s a bad idea to depend entirely on it, because since it has to be cheap, individual uses aren’t worth very much.

Supplementing this with ads is not acceptable, either, so it’s best if this is one of a couple ways to do it which you make available.

Consider using another of the ones further down, to make sure it’s all profitable and manageable.

Pay Per Month:

This is the middle of the road subscription model with a small span and manageable prices, but lets them out whenever they decide they need to. This is a short enough term for them to feel comfortable, but long enough to encourage loyalty with real history capture being available.

It also makes them worth a lot more individually.

Pay Per Year:

This is just a much bigger one, and it is really only fitting for enterprise solutions of a large scale, and in the case of these, the price tends to be quite higher here, and due to that size, it’s often justified at least on paper, but that’s difficult to really be sure of.

Having all three of these last ones available depending on a company’s needs, budget and scale is a very good idea, because it means you’re workable to all possible demographics without the struggle of freemium, which as I said, was never a model I was ever that fond of or confident in.

The Problem With Variance:

If it lets customers shape the size of their experience, and have many ways to get out sooner than you tend to want to allow, you do have that problem of inviting loyalty failures.

But, this is a risk you have to take, and it is worth it and effective. I’ll tell you why.

You see, if you show you’re not in need to trap the user for extended periods of time, you show that you feel confident that your design is good and useful enough that they’ll definitely want to stick around. In that case, why worry about the easy and early outs being made available?

But, most companies only use one of these SaaS business models at a time, or try traditional mercantile models based around traditional local on premise software sales practices. That just doesn’t work. So, pay close heed to these models, and think long and hard about which one works best for you, because choosing the wrong one for your demographic, industry or corporate culture will absolutely end in disaster.



Omri is the Head of Demand Generation, as well as the Lead Author & Editor of the SaaSAddict Blog. Omri established the SaaSAddict blog to create a source for news and discussion about some of the issues, challenges, news, and ideas relating to SaaS and cloud migration.