Jul 13, 2021
The Goldilocks Zone of SaaS Metrics
What’s an OK churn rate [conversion rate, ARPA,…]? Hint: It depends.
Founders often ask me questions like:
- What is an OK churn rate?
- How does our conversion rate compare to other startups?
- How are we doing in terms of CAC:LTV?
These are just three examples. What they show is that founders are keen to find out how their business is performing relative to their peers. This makes sense. Comparing various aspects of your business with those of other companies can help you find out the biggest areas for improvement.
If you know, for example, that your conversion rate is sub-par, it’ll be worth your time to understand why and how to improve. If your conversion rate is best in class, on the other hand, rather than trying to squeeze the last fraction of a percent out of your conversion funnel, you’ll probably want to focus on improving other areas of your business.
The challenge is that no single metric is very meaningful if considered in isolation, which is why my answer to these questions usually starts with “it depends”.
For example:
- A lower trial-to-paid conversion rate is OK if you’re attracting huge numbers of trials at a low cost.
- A higher churn rate is OK if you have a product with a built-in viral loop.
- A longer CAC payback time is OK if you have strong negative churn.
- High CACs for paid acquisition may be OK if your blended CACs are much lower.
Many of these variables are interconnected, so you always have to look at the entire picture. In some cases, two lower-level variables can act together to impact a higher-level variable. For example, if you’re a self-service SaaS company, your visitor-to-trial signup rate multiplied by your trial-to-paid conversion rate determines your total conversion rate from website visitors to paid customers, which is what ultimately matters.
In some of these cases, the two drivers ‘compete’: If you manage to increase your visitor-to-trial conversion rate, your trial-to-paid conversion might deteriorate because you get a larger number of less qualified trials.
What this simple chart illustrates is that a lower trial-to-paid conversion rate is acceptable if you have a higher visitor-to-trial conversion rate, and vice versa.
OK, so it’s hard to say what an acceptable trial-to-paid conversion is without looking at your visitor-to-trial conversion and your total funnel conversion rate. But what, then, is an acceptable total conversion rate? Well, that also depends!
It depends, for example, on:
- How much are you paying for a website visitor?
- What’s your ARPA?
- What’s your churn rate?
If you think this is a case of turtles all the way down, fear not. In the end it all comes down to LTV/CAC, which is why this is one of my favorite charts:
What you can see here is that if you have a high LTV (AKA you’re going after elephants or whales), you can afford high CACs. If you have a low LTV (AKA you’re selling to mice or rabbits) you better have low CACs — or you’ll end up in the graveyard quadrant of the chart. If you find yourself in that quadrant you’ll have to decrease your CACs and/or increase your LTV in order to get to the habitable zone of the chart.
There may be reasons to temporarily accept a lower LTV/CAC ratio, e.g. if you’re spending aggressively to get market share. And before you have strong Product/Market Fit you shouldn’t worry about LTV/CAC because it’s all about learning and getting to PMF.
But ultimately SaaS businesses make money by monetizing customers, over their lifetime, at a multiple of what it cost to acquire them. So you’ll have to make your LTV/CAC ratio work.
In conclusion, here are a few tips:
1) If you’re trying to find out what churn rate [or conversion rate or ARPA or … ] is acceptable for your business, start from LTV/CAC and work backwards from there.
2) As a rule of thumb, a good LTV/CAC ratio is 4. There’s no science behind that number, but if your LTV/CAC ratio is 4 it means that after spending 1x your CAC on CAC (duh) and 1–2x on G&A and R&D you have 1–2x to cover the costs of capital and generate profits, with some margin for error.
3) Keep in mind that your LTV for current or future customers is always an estimate based on the behavior of past customers. You won’t be able to measure your CAC with 100% precision, either, because of various kinds of tracking errors and attribution uncertainties. Therefore you have to work with simplifications, conservative assumptions and leave a large buffer for error.
The simple chart above showed that if you’re aiming for a 4% total conversion rate, you’ll have to land somewhere on or above the green curve. Similarly, you can imagine a 3-dimensional space with ARPA, churn rate, and CAC as the axes. Somewhere in that space is a body that represents a CAC/LTV ratio of 4 or higher. We can call that the Goldilocks Zone of SaaS Metrics¹⁾.
Now go find your place in that zone!
PS: If you read this post hoping to get a precise answer to the question “What’s an OK churn rate?”, I’m sorry. 😉 As a small consolation prize, here’s a somewhat older piece about SaaS metrics with churn benchmarking data from 1500 SaaS companies.
¹⁾ One might argue that the Goldilocks zone isn’t a good analogy because for a region in space to be habitable it can’t be too hot or too cold, while in the case of LTV/CAC, there is no “too high”. That may be true, but then again, if your LTV/CAC ratio is 20 it means that you’re not spending enough on customer acquisition.
Thank you, Nathan Benaich and Stefan Smalla, for reviewing a draft of this post and your valuable feedback!
