Churn Rate is one of the most important metrics you can calculate for your business. It’s one that investors and boards of directors stress constantly, but more importantly, it’s a signal of whether or not you have a viable business on your hands. And yet, finding out which Churn Rate formula to use can be surprisingly tricky.
A key part of the problem is that public companies don’t disclose metrics like Annual Recurring Revenue (ARR) or Churn Rate in their filings, so you can’t model your own metrics off of theirs. And attempting to work backward from what they do disclose gets iffy fast.
Despite the challenges of calculating churn, getting it wrong can have some painful upstream effects, leaving you with an incorrect understanding of metrics derived from churn like Customer Lifetime and Customer Lifetime Value. So with over forty different churn calculations at your disposal, how’s a CEO to decide how to calculate this information?
Quick Tips to Consider When Calculating Churn
Before we get into the details, there are a few things to consider that might just help with Churn Rate formula overwhelm.
- Align With Your Investors: If you present a number in a board meeting and investors don’t like your math, it’s likely to derail the entire meeting—even if you put a lot of thought and consideration into your calculations. So save yourself the headache and ask your investors if they prefer all of their portfolio companies use a certain calculation.
- Know the Difference Between Renewal Rates & Retention Rates: We’ll get into this in more detail soon. For now, remember that your customers likely have a mix of different contract lengths. If you only look at retention as the inverse of churn, you might be dramatically underestimating your customer churn problem.
- Calculate Gross and Net Revenue Churn: It’s not one or the other. It’s both. Make sure you keep tabs on both gross and net churn regularly and that you understand the differences between them.
- Use Cohorts: A cohort is simply a group of customers that are grouped together around a shared characteristic. Often it is used to group customers acquired over a given time period. Calculating churn with a blended cohort is a recipe for tears and gnashing of teeth.
Churn Rate Definition
How you define and calculate churn may vary depending on your business model. For simplicity’s sake, we’ll be talking about B2B SaaS businesses, where churn is the percentage rate at which customers cancel their recurring revenue subscriptions.
Your Retention Rate, the inverse, is simply 100% minus your Churn Rate. It’s an easy calculation…once you’ve found the Churn Rate.
There are three key ways to calculate churn: by customer count, by licensed seats, and by revenue. Let’s dig into each.
Churn Rate Formula by Customer Count
Calculating your Churn Rate by customer count is one of the most straightforward ways to look at this all-important metric, but it’s not the strongest for truly understanding the impact of churn on your B2B SaaS business. This is why I recommend only reviewing this calculation in comparison to a revenue-based churn calculation, which we’ll talk about shortly.
Customer Count Churn Rate = (Beginning Customer Count – End Customer Count) / Beginning Customer Count
As you can tell, this Churn Rate formula only looks at the total number of customers at the beginning and end of a period of time. For example, let’s say you’re calculating monthly churn between September and October. Using this formula, your Beginning Customer Count would include all customers acquired before September 30. For the End Customer Count, you’ll look at how many of those customers are left on October 31. You could use the same reasoning to look at churn on a quarterly or annual basis.
While it’s a helpful, easy-to-calculate way of looking at churn, this method essentially gives all customers equal weight in the eyes of your business, and though you may love all customers equally, they’re rarely worth the same in terms of revenue. The bigger the variation in account size, the less helpful this formula becomes.
Churn Rate Formula by Licensed Seats
Product leaders (especially those in multi-product companies) are particularly concerned with the licensed seats calculation because they can calculate individual product Churn Rates using it. It’s a more detailed level of information than CFOs, CEOs, or board members are likely to care about, but it can help those driving the product roadmap to understand where customers are getting the most—and least—value.
Licensed Seats Churn Rate = (Beginning Number of Users – Ending Number of Users) / Beginning Number of Users
Much like the customer count Churn Rate formula, this one looks at total users at the beginning and end of a period of time. The trick here is making sure that if multiple product team members are reporting on this metric that they’re consistent in how they calculate it and that they’re comparing the same period of time.
It’s worth noting that in this calculation “users” are a placeholder for however your product monetization model works. Some business it’s users/seats, for others it’s some other metric like monitored endpoints. Nevertheless, the concept holds true across monetization unit types.
Churn Rate Formula by Revenue
Every customer you have is worth a certain dollar amount in recurring revenue, whether MRR or ARR. If your ultimate goal is to understand the business impact of losing customers (and it really should be), revenue-based churn calculations are the way to go.
Unfortunately, they’re also the most complex, and you can look at it in terms of gross or net churn. Here’s a simple breakdown.
Consider Gross vs. Net Churn Rate
The key difference between gross and net churn is how expansion and contraction are treated. If one happy customer stays on and adds additional products (i.e. cross-sell) or users/seats (i.e. upsell) while an unhappy customer leaves, the additional revenue from the happy customer could mask some of the revenue lost. For that reason, it’s a pretty good idea to look at both.
Gross Churn does not take expansion into account. It only includes revenue lost due to cancellations. So any add-on or license expansion revenue should be tracked appropriately and broken out separately when calculating this metric.
Gross Churn (Monthly) = (Churned MRR in Period + Contraction MRR) / MRR at Beginning of Period
A handy way to remember the difference is “Gross is gross.” Gross Churn can only show you what you’ve lost, which is pretty gross, even if it’s necessary. One thing to note: Depending on how realistic (or depressing) you want your Gross Churn number to be, there are varying thoughts on whether or not to include downgrades in the number as well. Including Churn + Downgrade will give you the most accurate picture, though maybe not one you want to tell anybody about.
Net Churn, on the other hand, includes upsells and cross-sells, as well as downgrades.
Net Churn (Monthly) = (Churned MRR – Expansion MRR + Contraction MRR) / MRR at Beginning of Period
The Net Churn number can go negative if your Expansion MRR outstrips your Churned and Contraction MRR. This is a feature, not a bug. (i.e. a good thing).
The hardest part here is making sure your financial data is properly organized so that you can plug the right numbers into the churn formulas above.
Revenue Retention Rate
Often, when it comes to discussions of churn on a revenue basis, the terminology is usually in terms of retention instead of churn. Why? I don’t know, maybe CFO’s are more cheerful people? Maybe ‘The Street’ doesn’t like to be reminded that your customers leave you?
The formulas above are likely more academic than two more formulas you’ll hear more often: Gross Dollar Retention (GDR) and Net Dollar Retention (NDR). (You may also hear these terms with “Revenue” substituted for “Dollar” but they are generally equivalent.)
GDR = (Beginning ARR – Churn ARR – Contraction ARR) / Beginning ARR
NDR is one of the ‘holy grail’ metrics in the SaaS world. It shows the total ARR you keep after expansions, contractions, and churn have been considered.
NDR = (Beginning ARR + Expansion ARR – Contraction ARR – Churned ARR) / Beginning ARR
Or more simply:
NDR = Ending ARR / Beginning ARR
If you want to have an exciting, investable business, your NDR needs to be above 100%, meaning that your existing customers are expanding their business with you over time. Investors will often expect a nice delta between GDR and NDR as a marker of a growing company.
Where Churn Rate Formulas Get Messy—And Errors Are Introduced
While the formulas are straightforward enough, in practice, it’s usually not as simple as plugging numbers into an algebraic formula. It’s important to think through some different criteria to make sure that the numbers you are getting are useful and not just “garbage in, garbage out.”
Churn Rate calculations commonly go wrong when you have any of the following conditions:
- Too Few Customers
- Extremely High or Low Churn Rates
- Extremely High Growth Rates
- Mixed Contract Renewal Periods
- Mixed Customer Types
How to Address Churn Rate Calculation Concerns
Now, you want to avoid all of these pitfalls when you start to plug numbers into your formulas. So consider the following a short checklist to run though when you calculate churn metrics.
Uniform Customer Population
When calculating churn, it’s important to compare apples to apples and oranges to oranges. If you have very distinct customer profiles/personas, pricing tiers, or value propositions, it makes finding trends in your data much more complicated. Define any major customer groups and isolate them as much as you can.
Matched Timeframe
Remember that Renewal Rate and Retention Rate are not interchangeable terms.
- Retention Rate: The inverse of churn (100% – Churn Rate), or how many total customers you’ve kept.
Retention Rate = Customers at End of Period / Customers at Beginning of Period
- Renewal Rate: A more specific rate that shows how many customers were retained, given how many were Available to Renew (ATR).
Renewal Rate = Customers That Renewed / Customers With Contracts Up for Renewal
For example, imagine that your customers have two agreement options: one-year contracts or two-year contracts. Now, for simplicity’s sake, let’s imagine you signed all of your customers at the same time, exactly one year ago, and they’re evenly split between one and two-year contracts. If 50% of your one-year contracts churn and you calculate churn using the entire customer population, your overall Churn Rate will look like it’s 25%. But in reality, 50% of the people who could churn did. And this spells major trouble for your business next year.
This is why it’s so important to look at ATR contracts when looking at churn and retention. Without breaking out ATR, your numbers will be artificially inflated and you won’t know the scale of the problem to address it.
Note that the above formulas were given for customer retention and renewal rates, but the same can be done for license or revenue-base retention and renewal rates.
Calculation Time Interval
In every formula above, there’s a “beginning” number and an “ending” number because we’re looking at churn that took place over a period of time. You get to choose the period of time you use, whether it’s a month, quarter, or year depending on your business.
So which time period should you go with? Here are some things to look out for.
Long Enough to Give You Data
If the period you choose is too short and you have very few data points, your churn metrics will swing wildly from one period to another. This can create a lot of stress in the company, as managers forget their college statistics and run fire drills to fix what looks like poor performance instead of what it is: noise.
Mixed Contract Lengths
If you have a combination of monthly, annual, and/or multi-year contracts, you have a few options.
The easiest option is to average out the contract renewal period. The formula would look like this:
Renewal Rate = Number of Renewed Customers / (Avg. Contract Renewal Period x Number of Customers Up for Renewals)
The harder, but more accurate way to deal with multiple contract periods is to calculate renewals using only the data from customers that were ATR.
Base Churn Rate
Many businesses like to have a “Base Churn Rate” that they can calculate regularly in order to keep a close eye on churn. This is a great idea, as it can help you address problems as you notice them and not after they’ve had time to grow.
A common mistake occurs when translating this Base Churn Rates into other periods. Businesses will try to convert their Monthly Churn Rates into Annual Churn Rates with simple multiplication. For example, if their Base Churn Rate was 2% each month, multiplying that by 12 months would show an Annual Churn Rate of approximately 24%, meaning an Annual Retention Rate of 76% (100%-24%).
If Churn Rates are very low, this simple math is often “close enough.” But once monthly Churn Rates cross 4%, the error starts to grow significantly as shown in the chart below.
The true way to annualize retention rates using monthly Churn Rates is not to multiply the rate by twelve months. Rather, use the following formula:
Annual Retention Rate = (1-Monthly Churn Rate)^12
Go Forth & Calculate Churn Metrics
As you’ve probably figured out from everything I’ve written, calculating churn correctly is complicated. And the reason is because companies vary so widely from one another in what they sell, how they handle contracts, and how many customers they have at a given stage.
Don’t forget that the ultimate goal here is not to get the answer right on a math test and move on. The goal is to end up with numbers that mean something to your business and that you can use to improve it.
The hardest part to finding the right Churn Rate formula is simply getting started. Once you’ve settled on the calculations that are most relevant to your business, it’s about plugging in the right numbers, checking in at the right intervals, and recognizing patterns.
Frequently Asked Questions About Churn Metrics
Should I include those who have canceled their subscriptions, but their subscriptions have not ended yet in my churn calculation?
There are differing views on this, but mine is yes, you absolutely should. Some folks think that you should use this as a signal to go “save” the customer with fancy retention plans (or software they’re selling). I think it’s mostly a bad ROI and use of resources. You’ve already failed that person. Can you save them? Maybe? Is it better to take the lesson learned and improve that part of your business so it doesn’t happen again? Absolutely.
Should you include customers that have “paused” their subscriptions when calculating churn?
Yes. Again, there are differing views on this, but a customer who stopped paying you is no longer getting value from your product. That is worth paying attention too, no matter the reason.
How should I account for “unavoidable” churn?
How do you account for unicorns crossing the road in your daily commute? Exactly, you don’t. Same answer, because unavoidable churn does not exist. This point will likely be a blog post in itself down the line.
How should I account for voluntary vs. involuntary churn?
Honestly, I doubt you will know the difference. It’s not worth worrying about. See previous point.
How do I account for growth within a month?
For B2B companies, it’s usually impossible for a customer to be less than monthly. Effectively, you can consider all the customers acquired within the month as the same cohort. Or in other words, everyone is considered acquired on the last day of the period. This allows you to not have to worry about growth within the month.
How do you calculate churn for a SaaS application with multiple price levels? How about with customers at significantly different spend (ARR)?
Make sure you’re following one of the previously stated rules when looking at churn, having a uniform customer population. If you have significantly different spend between customers, segment the customers into groups based upon ARR.
Should I include free or trial accounts in my Churn Rate formula?
No. A customer is a customer. This would break uniform metrics. They haven’t paid you anything, so they are at a different value proposition and very likely a different persona.
What are some good analytics products for SaaS businesses that help with customer churn, segmentation, and other product metrics?
I will cover this in a separate post. But hint: It does not include Google Analytics.