A user is someone that uses your SaaS product, right? Or is it someone that signed-up? Or someone that’s active? Or someone that logged in a few times? Hmm.
Okay, so maybe defining a user is hard, but defining a customer is easy, right?
A customer is someone that pays you for your product or service. Even if they’re still within the legal timeframe for a refund? Or a contractual “cooling off” period? Or if they’re within the 90-day “stick point” (if they make it past 90-days they’ll stick around for a long time)? Or…
Wow, so even defining a customer isn’t as straightforward as it might have seemed.
And it gets even messier if you’re in a market with a more transient customer base (i.e. the level of real unavoidable churn is high), if you offer a completely free or freemium product, if you just launched with a lot of early-adopter interest (i.e. the “Product Hunt effect”), etc.
To get honest about what’s going on in your company, you need to modify the customer (or user) definition, which is the main input into how you calculate the core metrics of your SaaS business.
Let’s explore this a bit…
Standard User Definition
You could totally just use the basic definition of customers and users for your calculations. If your board or investors want to see roll-up numbers, by all means, use roll-up numbers (most good boards and investors will want to get more granular, however).
But, just like all high-level roll-up metrics, the real inner-workings of your business can be hidden (obfuscated – often unintentionally… sometimes not) and the actionability of those metrics will be limited.
So I’m not trying to talk you into anything here, but if understanding what’s really going on in your business – and then being able to do something about it – are useful to you, this might be something to consider. Cool.
Defining Users Realistically
I’ve covered the topic of Free Trial users being a vanity metric before, and I just recently covered the topic of “active users” being a vanity metric, as well.
Ultimately, when it comes to defining “users” you probably want to start only with those that are actually “engaged” with your product or service (whatever “engaged” means… hopefully it’s well-defined in your world).
This means getting away from low-value metrics like “signups” or “installs” or “logins” or even general “activity” and into specific metrics like “contextual activity” or activity that indicates whether the user is doing something from which they will derive value.
I’ll be honest… this will likely reduce the number of “users” you have – which will cause a hit to the ego – but it will give you a better, more realistic view of what’s really going on in your business.
And of course, the same thing goes for customers…
Not All Customers Can Churn (Right Now)
In the Customer Success world, we often talk about calculating churn for a given period of time only against customers that could churn, i.e. those with an expired contract, who are up for renewal, with an out clause, etc.
Those that are otherwise bound by a contract couldn’t churn this month, so if we figure them into our churn calculation, we might look like we’re doing better than we actually are.
I’m no mathamagician, so I’ll keep it super-simplified, but here’s how this plays out (for these simple examples I’m talking about customer or logo churn and not revenue churn):
The Standard Churn Calculation (Super-Simplified)
Started with this many Customers: 100
This many Customers Churned: 10
Left with this many Customers: 90
Customer Churn Rate: 10%
Customer Retention Rate: 90%
Only Counting “Churnable” Customers
Started with this many Customers: 100
This many customers are in the “unable to churn” cohort: 13
Started with this many Churnable Customers: 87
This many Customers Churned: 10
Left with this many Actual Customers: 77
Customer Churn Rate: 11%
Customer Retention Rate: 89%
While this example might not look like a big deal – just a 1% difference – it’s actually a 10% increase in churn rate if we look at the customers that can churn. Rest assured, that small percentage will quickly compound into something significant.
Okay, so there are customers that aren’t eligible for churn, that makes sense, but to get an even more realistic metric, we need to acknowledge that there customers that we really shouldn’t consider customers… yet.
Not All Customers are Customers (Yet)
In the same vein as only considering customers eligible to churn in your churn calculations, you should only count customers that are actually customers in all of your other SaaS metric calculations.
Many SaaS companies will have a cohort of customers that – immediately after the sale or right after their conversion from free trial to paid customer – really shouldn’t be considered a customer, yet.
Maybe it’s a cohort of customers that came from being featured on Product Hunt, or from being included in a bundle with other SaaS products, or by running a discount campaign (the way I do discounts doesn’t cause this problem, but I digress) where we aren’t sure if they’ll stick around.
Or – like I said at the top – maybe it’s a cohort that’s still within the legal timeframe for a refund or they haven’t reached the “stick point” for your product – so we don’t want to figure them into long-term revenue projections just yet.
We also don’t want to figure them into “real” churn numbers yet because, for this “too-early” cohort, their churn rate may be higher. Also, their churn REASONS will be different from more tenured customers (were oversold, onboarding didn’t happen, they exercised an out clause, etc.). Of course, you’re going to work to reduce the churn rate in this cohort just as you’re doing for the rest of the customers, but you’re just not figuring it into the overall churn numbers.
For instance, in low-touch, self-service, quick configuration products, I suggest that you wait one or two billing cycles (60-90 days) post-conversion before I say they’re a real customer. This is doubly important if you do anything, err… shady… like forced continuity or if you require a credit card to start your free trial; after 2-3 billing cycles, it’s less likely they just forgot to cancel.
Again, I’m no mathlete, but it’s actually not that hard to come up with some pretty significant results:
The Standard Churn Calculation (Super-Simplified)
Started with this many Customers: 100
This many Customers Churned: 10
Left with this many Customers: 90
Customer Churn Rate: 10%
Customer Retention Rate: 90%
Including “Churnable” & Excluding “Too-Early” Customers
Started this time period with this many Customers: 100
This many customers are in the “unable to churn” cohort: 13
Started with this many Churnable Customers: 87
This many customers are in the “too early to be considered” cohort: 17
Started with this many Actual Customers: 70
This many Customers Churned: 10
Left with this many Actual Customers: 60
Customer Churn Rate: 14.3%
Customer Retention Rate: 85.7%
As you can see, by excluding customers that are unable to churn and those that are in the “too early” cohort, we have a real churn rate of 14.3%… or nearly 50% greater than what our “conventional” churn rate is. That’s significant.
And you’ll see similar differences in other key SaaS metrics – like Customer Acquisition Cost (CAC) and Monthly or Annual Run Rate – when you start calculating based on more meaningful definitions of users and customers.
Of course, you can use whatever metric you want externally to show how awesome you’re doing, but internally – if you want to see how things are REALLY going – you should be honest in how you measure things by starting with accurate definitions of customers and users.
About Lincoln Murphy
