SaaS Math: Activation, Retention and Churn

So far in the SaaS Math series, we have looked at why investors love SaaS, definitions of the key terms in SaaS math and last week we looked at user acquisition math. This week we’re talking about activation, retention and its polar opposite – churn. As you will see, churn and retention are the most important metrics to the long-term success of a SaaS business.

To recap, activation is just getting that 1st use of your product or service. Churn is the % of users / customers that abandon the service over time. This can be measured weekly, monthly, quarterly, etc. You will want to measure churn for users and churn for customers (assuming you have a free trial or freemium product). Retention is the inverse  – the % of users that you keep.

Let’s look at an example of a freemium business. It has a free base version and converts a portion of its users to a premium service. In this case, you need to look at churn of both paid and free users.

The beauty and the challenge of freemium is that in the absence of having to pay money there are no barriers to users trying you out. Similarly, there’s nothing keeping them around either (other than an amazing experience and utility). Easy come, easy go.

Many entrepreneurs ask when they should consider a free user to have churned out? After all, they could theoretically come back any time. To be clear a user account is never deleted since that can mess up the database, but I usually apply the telco operator standard measure of churn. If a user has not logged in for 90 days, then you should consider that they’re gone forever and move their account to an inactive status.

Going back to our example. Here is a simple illustration of user acquisition and activation for this freemium business.

You can see here that:

New users are growing at 10% a month.

30% of new users use the service at least once.

5% of new users convert to a paid account.

A simple churn calculation suggests that 7% of customers cancel their paid accounts each month. This suggests that the average customer keeps her subscription for about 14 months (1/ 7%).

To truly understand retention and churn however, you can’t just look at this high level stuff. You need to go deeper down and do a cohort analysis. See, users don’t just drop off at a steady rate each month. With free users you usually see the following pattern:

– Most never use the service. Maybe they just signed up to see what it looks like or because they were invited. Who knows?

– Many will try the service once or twice then move on. So the 30 day churn rate is usually very high.

– 60 and 90 churn rates are also high (though not as high as 30 day)

– If you can keep a user past 90 days, they usually stick around for a long time.

You see similar patterns with paying customers. If you offer a free trial (say 30 days) then bill after, I recommend not calculating churn until you have been through one successful billing cycle. Many free trial users cancel their accounts during the trial period or as soon as you actually charge them.

So digging deeper into paid customer churn:

Here, 30% of customers cancel after the 1st month. After 2 months, 45% have canceled. The balance of the cancellations in any given month come from older cohorts.

This example only shows partial data to keep things simple, but if you do deep cohorts by month for active and paying users you will get much more accurate data on how long users & customers stick around.

Churn – the most important metric to a SaaS business

Finally, the most important metric for the long term success of a SaaS business is the customer churn rate. Why? Because the lower the churn, the more months subscription revenue you generate from each customer. This increases customer lifetime value, enables you to invest more in customer acquisition, which makes you grow faster which increases your valuation multiples! So, reducing churn creates this virtuous cycle of benefits to your business.

In this simple example you can see the importance of churn reduction. A 50% reduction in churn results in almost a 3x increase in the lifetime value of a customer.

In my experience, you need churn rates to be 5% / month or lower in order to have a true long term money maker on your hands.

  • http://www.guaranteedppc.com Guaranteed PPC

    Hey, Mark! I bumped into this article and I enjoyed reading it. :) Thank you. I also wrote a piece on churn rates and customer reactivation that you might be interested in, which also gives you my perspective on it.

    You can see it here: https://guaranteedppc.com/customer-reactivation-using-ppc-your-money-making-how-to-guide/

    Thanks again!

  • Eyal Toledano

    Hey Mark,

    I have trouble agreeing with 5% monthly churn being acceptable as it does translate to ~46% churn annually which is a disaster…

    I think 5% ANNUALLY is acceptable. What do tou think?

    • http://startupcfo.ca/ Mark MacLeod

      Hey Eyal,

      5% is not a target. In my experience, you can’t get positive unit economics above this number. 5% annual is just not realistic

  • http://twitter.com/undefined @undefined

    Mark- this is a great article. One of the things we like to look at is actually the relationship between LTV, CAC and rate of CAC recovery. Something I call the "Tolkien Number"- that is one SaaS metric to rule them all: http://www.sethlieberman.com/saas-math-calculatin

  • Pingback: Churn Rate – The most important metric for a monthly revenue business | Get Personal()

  • Pingback: 3 Invaluable Resources For Learning About SaaS Freemium Pricing and Metrics | | SAAS TimesSAAS Times()

  • Stephen

    What is more common? Calculating churn based on static point at beginning of year or based on number at the start of a given month?

  • http://www.cloudnco.com Matthieu Dejardins

    Thanks Mark, your perspective on the "normality" of these KPIs is very interesting there…
    The Churn is the equivalent of the bounce rate in Web Analytics with more immediate consequences :)

    I agree with your Fremium recommendation, if you look at all the big players, to build massive product usage, this is the ideal way: Linkedin.com, Box.net, Dropbox, Spotify, Pandora, Hulu, Slideshare, Google Analytics, Google Drive, etc.

    Even if you are converting a fraction to Paid use (i.e. 3 to 10%), your equation is on so high volumes that is economically viable!

    Regarding the tool, Anne Holland a very useful financial model available here: http://www.subscriptionsiteinsider.com/public/log
    It is for subscription sites but perfectly applicable to any SaaS or Cloud Services.