The Economics of Micro SaaS

Not all SaaS companies have equal margins. Today I wanted to go through each of our products and talk through net margins. It's useful to get a sense of what  margins could look like. I also threw in a small productized service to compare it against.

Screenshot API - SaaS

Screenshot API did ~$1,250 last month. We've tripled it since we bought it, but this product in particular has high operating costs (servers). It cost us around $440 last month to run the service.

The problem is that taking these screenshots takes a lot of (cpu) time, and since we have it set up to autoscale, our costs increase with the number of requests.

We didn't catch this in diligence because through diligence we failed to understand some of the architectural shortcomings of the product.

Also, our primary growth channel for screenshot is google ads. This is also (obviously) expensive, but is actually still revenue positive for us. We can pull this lever up or down to help with costs but until we find a better channel, this one is going to stay.

So back to the numbers:

• Topline: $1,250
• Servers: $440
• Marketing: $250
• Net: ~$560
• Profit Margin: 44.8%

When you hear SaaS, you typically think 70-90% gross margins. That's not always the case. In this instance, we have particularly high costs to perform the service. That's both a technical problem (we can solve it with code) and also a business problem.

Toybox - SaaS

Toybox has been technically challenging for us. The codebase we bought was falling over itself and we've had to drop a lot of time and cash into fixing that. When we bought it, there were close to $2k in server costs hosting on Heroku (don't get me started on Heroku... it's like crack, feels good for a little while, but ultimately will kill you). We just moved it to Google and lo and behold looks like we'll be under $300 in hosting costs. A huge improvement. With the new server costs in mind let's take a look at the numbers.

• Topline: ~$6000
• Servers: $300
• Marketing: $0 - working on this!
• Net: ~$5500
• Profit Margin: 91.66%

Low 90's feels about right for this one. There are a lot of users, and a lot of rich media (images, etc) but ultimately the product is a simple bug reporting and feedback tool that isn't doing anything crazy on the back end.

We haven't figured out a decent marketing channel yet for toybox. We're a small team and generally only get to focus on one thing per company at a time, and since engineering was the biggest problem, we haven't really looked at growth yet. We'll be doing some experiments with Cold Email this month and can hopefully find some messaging that resonates.

Sheet.Best - SaaS

Sheet best is awesome. I keep using it for personal projects and it's so simple I kick myself for not having thought of this idea first... that's alright, we bought it instead.

Sheet best is growing like a week. We get ~100 new users a day. So far we've doubled it and really haven't done much in the way of marketing (lucky).

Let's take a look at the numbers:

• Topline: ~$1,200
• Servers: $40
• Marketing: $0 - working on this!
• Net: ~$1150
• Profit Margin: 95.83%

We've gotten lucky so far. This product gets written about and grows organically. It's the best, but it also might run out at any time and we're not prepared for solid outbound for it yet. Also the servers could use a little beefing up before we'd be ready to double again, but for now sheet best just keeps chugging along.

Cold Email Studio - Productized Service

I also have a little marketing agency. Let's take a look at those numbers just to put things into perspective:

• Topline: ~$20,150
• Servers: $0
• Marketing: $0
• Other Costs: $500
• People: $12000
• Net: ~$7650
• Profit Margin: 37.9%

A few things could be misleading here. Our topline includes our regular MRR plus we sold a quarterly plan so actually that money can't all be recognized as profit since I have to pay wages against it in the future. We do our own outbound stuff so that's zero, but still, that cost is baked in to our people cost.

There are plenty of industries that would kill for a ~40% margin. But compared to even a 'Bad' SaaS, the margins are still low. This doesn't mean it's a bad business. It's a great little business that I have no doubt will reach $1M ARR next year.

I wanted to put this in here to give you some real numbers against real companies. It's not necessarily a bad thing to have lower margins. You might, for example, be able to make it up in volume. Different margins just mean a different kind of business.

Conclusion

Profit margins even amongst SaaS companies can be wildly different. I often point friends to read The New Business of AI to see just how hard it can be to build an AI SaaS with margins that look similar to 'traditional' SaaS apps (like Toybox). That's been my experience too, my last co was a machine learning SaaS that had much lower than 90% margins. Functionally, the margins looked much more like a service company than a SaaS company.

Overall though the margin doesn't matter that much without the context of volume. Low margin high volume can be = to high margin low volume. So take these numbers with a grain of salt, but it's always nice to look at real numbers just to be aware of what kind of business you actually have.