Churn is one of the most important metrics for your SaaS application. Whether your app’s revenue curve roars upwards like a Falcon 9 or whether it pancakes into the Valley of Despair like a Sumo-ringer splash diving – it all comes down to your churn rate.
There are a dozen ways to calculate churn, a dozen reasons for churn and hundreds of ways to fight churn; entirely too much for today, so I am going to focus on a type that gets next to no attention and which I call „onboarding churn“.
The quick and dirty basics
Churn is defined as „how many users out of 100 who start as customers at the beginning of a month cancel out and are no longer your customer at the end of the month“.
For example, if you have 200 customers at the beginning of August and 30 of those cancel any time between August 1st and August 31st, then you have a churn rate of 15%.
You can calculate churn either by looking at your users or by looking at the revenue they represent. You usually want to look at the revenue-based churn rate as it is more significant for your business. After all, losing two users on the $99/mo is a harder hit than losing ten users at $5/mo.
15% churn is not uncommon for early stage startups. You don’t have product/market-fit yet, you’re likely missing some essential features, your onboarding is far from perfect and so on. This all contributes to high churn rates.
Looking at churn in more detail
Having that number – commonly called „aggregate churn“ – is better than nothing, but it doesn’t exactly give you details. Sure, 15% churn is high, but it doesn’t tell you how to solve it.
Often the users on your lower-tier plans are more likely to churn, and people on the expensive plans are less likely to churn. Talk about counter-intuitive.
For my product, I see the same thing: users on the lowest plan churn out very quickly, whereas the people on the highest plan are with LinksSpy the longest. You don’t get that information from looking at your aggregate churn.
Now, a completely different question is: When do users cancel? Surely, their reasons for cancelling are quite different when they cancel after one month vs. when they cancel after five years. But you don’t glean that information from aggregate churn either.
Again, I have the same problem with LinksSpy. The longer someone is with me, the less likely they are to churn. The churn rate in the first 3 months is 44,1% and after that drops to about 7%.
There are a lot of different ways you can slice up those 15%, but let’s focus on only one type: onboarding churn.
The Strange Case of Onboarding Churn
Onboarding churn is a term I newly coined, so I sure hope it sticks – Mainly because I want to write a book and lecture about it and become a multi-gazillionaire. In the meantime my definition of onboarding churn is churn that happens in the first 2-3 months of a user’s lifetime, because users either don’t get value from your app OR they don’t see the value they get.
I already mentioned that LinksSpy sees quite a bit of onboarding churn, but there are a lot of other companies that face the same value. For example, Moz has onboarding churn of – low and behold – 40%.
In fact, Sarah Bird of Moz was the first to introduce me to the concept of onboarding churn. When she mentioned once that they calculate churn only after a user has been paying for three months, I literally thought „What a cheap trick“. Only after I launched my own product and facing the same problem did I realize that there is some merit to that approach. Kudos, Sarah, for teaching me that lesson before I was ready and my sincerest apologies for thinking you were somehow „wrong“.
As it is there are good reasons to think of onboarding churn as yet another step in your customer acquisition channel. After all users are not really on board as long as they do not use your product.
Free Trials do NOT Prevent Onboarding Churn
The thing that is – euphemistically put – interesting about onboarding churn is that most SaaS/subscription applications have a free trial. So why don’t people figure out whether they like the software during the free trial and cancel before they are charged? After all, that’s what trials are for!
I don’t have an exact answer to that, but I guess that people see initial value in the app and promise themselves to „look into it in a week when there is less on my plate“. Fast-forward a week and they still don’t have time – but next week they will; and so on and so forth. Yeah, that’s the best explanation I can come up with.
Free trials don’t prevent onboarding churn – at least not all of it. Look at Moz! Their trial is 30 days and they still have onboarding churn rates of 40 percent.
In-App Onboarding Does NOT Prevent Onboarding Churn
Surely, having those fancy in-app tutorials that guide you through using the app with those nice bubbles help? Yes… somewhat, they can help if you get them just right.
I’ve revamped the in-app tutorials for LinksSpy at least three times by now, constantly improving on customer feedback. The latest version will go live with the relaunch and will focus on a completely new set of features.
How To Nip Onboarding Churn in the Bud
First of all, you can’t entirely get rid of onboarding churn. Some percentage of users will always sign up for your app, start paying you, never use it and then cancel. Expect the lower bound to be around 20-30%. You can probably drive that number down by going to extreme measures (My friend suggested flying out to them, taking their kids to soccer practice and cooking dinner). But most of those things are not justified for a <$1,000 LTV product.
So you (and I) have to live with a certain amount of onboarding churn.
However, what you can do is to dramatically reduce your onboarding churn.
Two years ago, I freelanced for an 8-figure SaaS company. They are selling access to semi-raw data. We came up with an A/B-test to test the following hypothesis: Delivering actionable advice – based on the data – each month will decrease 30/60-day churn.
We did a minimal test; sending just one email in the first month. The test group completely crushed the control group. Sending just one email decreased their 30 day churn rate by 40%.
It meant that they made an ROI of some 400% on what I charged. From running the split test. For one month. With only half the cohort receiving treatment. Ignoring prolonged customer lifetime.
That’s the power of reducing churn for a company at scale.
Use Emails to Reduce Onboarding Churn
So, emails can work fantastically well. Here are a few ideas what to send:
- reminders to use sticky features
- demonstrations of received value (e.g. „You saved $192 this month using our software“) a.k.a. „Get our users promoted“ emails
- reminders to use the app after inactivity
Better First Run Experience
Secondly, having a better onboarding/first-run experience can decrease onboarding churn.
- Never let your users hit an empty page. If you don’t have data yet, show fake data
- Lead users down the Minimum Path to Awesome in your product. Eliminate distractions.
I can give you an example of the first point. This is what LinksSpy customers see when they enter the application:
See all that beautiful white space that designers drool about? Customers don’t like it. Your churn rate hates it.
Now here’s what they will see after the relaunch in 1.5 weeks:
Look at that! Fake data and an in-app tour! Beauty.
Get to Product/Market Fit
Furthermore, reaching product/market fit will lower your churn rate across the board. This is not an easy task. Even after you have found a problem and your product presents a viable solution to the problem, you still need to figure out how to reach your target market.
Rob Walling had this problem in parts with Drip: Drip was initially a tool to collect email addresses and send autoresponders.
He got quite a few customers with this tool, but the churn rate was rather high. Only after he pivoted into Lightweight Marketing Automation did the churn rate go down.
Subsequently, the product and MRR blew way up. The huge problem here is that you need to make a decision to either build new features to satisfy your current audience OR pivot to another audience with the same product. It’s hard to know which steps to take.
Software With a Service
Lastly, you can improve your retention by offering a Done-For-You (DFY) service on top of your product. Users don’t want to learn how to use your product. They care for the results.
This is exactly the path that LinksSpy is going to take in the upcoming months. We will discontinue the $19/mo plan and instead introduce a $499/mo plan where we manage the whole outreach marketing process for your website. It’s completely hands-off and you will get a report at the end of the month detailing the results we got you.
To summarize all this: Onboarding churn is a thing. You will lose most of your customers during the first 3 months and churn rates do drop after that. This effect is caused by a number of things, most notably by not getting your users invested in your product.
Ways out of high onboarding churn are product/market fit, better first-run experiences, email (retention) marketing, and offering a Done-For-You option for your product.
Let me know your thoughts on the topic. Is there anything you struggled with? How did you overcome that challenge?