Prevent Churn: An Easy Guide for All
How to prevent and reduce churn?
I’m a true believer in that churn starts day 1. From the first minutes with your product, you have only a limited timeframe to educate your customers on your value proposition, how the product works, get them set up, and help users adopt the product.
The moment a customer buys your product means only that a potential customer trusts you enough that your solution will help them accomplish their goals. It doesn’t mean it will help, and it doesn’t mean the customer will keep trusting you.
This is also why preventing churn shouldn’t start with customers canceling your product — it’s extremely hard to win them back then. They already lost trust. So preventing churn starts with day 1.
I can tell you that more churn occurs in the first 3 months of the customer lifecycle than in any other period.
– Corey Haines, https://www.swipefiles.com/articles/ltv-cac-ratio
Here’s a telling, scary, and motivating picture when you see it for the first time:
X axis represents the number of days any user performed an action on the customer account in the last month. Y axis represents the number of customers — their accounts.
Left side chart shows paying customers that did not set up your product. There’s no activity for most of them. No activity equals churn will come sooner than later. “Patient is basically dead” meme comes to mind.
Right side chart shows paying customers that did set up. Now look at these beautiful large green bars within 21–31 days of activity buckets. And look at these small bars at 0–5 days buckets.
Lack of users activity, that also indicates lack of long-term retention of users, will make or break your SaaS business. If only there was a way to identify customers at different stages of setting up and adopting your product, and learning how to help them out…
Introducing you to the churn prevention funnel!
Churn prevention funnel
I made up this name — Churn Prevention Funnel, so let me explain what it means for me. Churn Prevention Funnel is a way to look at a cohort of paying customer accounts that allows you to:
- identify customer setup rate and customer adoption rate
- Identify well targeted segments of at-risk customers before churn happens.
- Identify why customers are about to churn, before churn happens
- Form a strong hypothesis on how to change the churn trajectory of at-risk segments of customers.
- And learn by the way why happy customers are happy, and be able to identify them, as they are great candidates for expansion.
Definitions:
Account — customer instance of the product that can have multiple users. Example: as a company when you buy SaaS products like Attlassian, Productboard, Miro, Figma etc., you purchase an instance of the product where you can invite developers, product managers, collaborators, designers, and others.
User — a person invited to an account. Typically users have to confirm their invitation.
Setup rate — percentage of paying accounts or users that performed all initial setup actions. Those can be multiple actions identified as a mandatory for an account to get value out of the product. E.g. for a CRM product one of such actions can be bringing data to it, or mapping your sales processes. User setup might mean e.g. that a person connects to their business calendar, and owns, or brought some customer data.
Adoption rate — percentage of paying accounts with users that are performing actions indicating getting value from the product. For Miro, or Figma, which are highly collaborative products, a proxy metric of getting value could be the amount of boards shared and viewed by invited people.
Churn Prevention Funnel example
Let’s say we’re building HealthyRemoter® — a remote time tracking app that truly promotes remote employees work-life balance. Take a look at how Churn Prevention Funnel might look like for a hypothetical app, in spreadsheet:
Add alt text
Prerequisites to build your first Churn Prevention Funnel
You’ll need:
- A little bit of patience
- Access to raw customer interactions data that you can transform to your liking. Any data warehouse solution should do, like BigQuery. This is to be able to identify key user actions connected with a specific account, and user.
- Friendly people around you like product managers, researchers, designers, business data analysts that will help you define key account and user setup actions, and core user actions.
- Friendly data engineer that will help you blend the data you need
- Spreadsheet to surface and filter this data
What about existing solutions tracking customer retention and churn? There are a handful of products that should be able to do that for you — receive user interactions in a meaningful way, and provide necessary user segmentation. I believe Amplitude, Totango, or Mixpanel could be your evaluation shots.
But hey, you’re still here, so you want to try & learn what you really need first.
The most important thing out of this whole exercise is to make sure you can actually close the loop and take cross-department (ideally), or within your team (good start) actions based on the insights you’ll get.
And if you’ll evaluate helpful software afterwards — it will be a more meaningful choice.
Step by step guide to build Churn Prevention Funnel
Step 1 — Define measures
Our measures for the Churn Prevention Funnel for our hypothetical HealthyRemoter® product could end up being defined like this:
- Key account setup actions: Admin invited 1+ user
- Key user setup actions: Confirmed user installed time-tracking app
- Core user actions: Confirmed user captured work time AND took 1+ recommended pause
Let me know in the comments, if you’d like to learn more about how to identify such actions using quantitative and qualitative research methods.
Step 2 — Make sure you have correct data instrumentation in place
Define your data model that allows you to run simplistic queries, and get results in the simplest possible form.
Pseudo SQL I like to optimize for:
SELECT * from Churn Prevention Funnel
WHERE First payment in Q3 2022 AND Data Snapshot From today
Here’s a sample result table moved to spreadsheet, with coloured data points that will get included in the funnel:
Step 3 — turn raw data into dashboard
Prepare formulas that will count accounts if they meet your measures criteria. Based on sample 8 accounts data from above, here’s an output Churn Prevention Funnel:
Step 4 — interpret and take action!
In real life the numbers of new paying accounts are typically greater. So imagine the top of the paying customers funnel doesn’t start with 8 accounts, but for example 1000 new paying accounts in Q3 2022.
Let’s go back to the first screen:
It quickly becomes obvious that for 750 of these accounts, users are not adopting the product.
They’ll churn first.
750 of 1000 accounts. That you have paid for to acquire.
What a pain!
Worry not! At least now you know.
Things you and your organization can do from now on
You can now immediately start talking with customers. As they have not churned yet, it’s far easier to catch them. In our sample case I’d try to understand:
- Why are confirmed users not installing the HealthyRemoter® app? Perhaps the issue is in confirmation rates? If you’ll look at the raw data, you’ll see confirmation rate is not the issue
- Why are users not capturing work time, or taking pauses, or both? Is this something in our control?
- For users that started tracking time, and are taking recommended pauses, what makes them lose the habit?
Soon enough, you should be armed with multiple opportunities to change churn trajectory.
You and your organization should also:
- Research specific segments of at-risk customers
- Feed this data into your CRM, customer experience, or customer success platform — place where Customer Success managers, Sales, and Advocacy can understand and make use of customer health score
- Analyze if the actions you take to prevent churn, made an impact in following account cohorts
FAQ
Do I need data snapshots to be captured everyday to get started?
No. You can take one data snapshot for a start. Then once a month. Then once a week. Manually. Then automate if needed. The importance lies in taking action, not live dashboards.
What if I don’t have a tool to visualize the funnel?
Stop worrying about visualizations, rather worry about your churning customers. Excel, Google spreadsheet, Airtable — use whatever is simpler. Once you have raw data, you can do calculations almost on a piece of paper.
Should I bring more data to the raw data table if it costs “nothing”?
Keep it as small as possible to focus on what matters. This table’s purpose is NOT to track user engagement in dozens of product areas. It’s to clearly paint a picture around core adoption drivers and reduce churn.
Thank you
If you find this article interesting, share it with your colleagues that focus on reducing churn.
This article first appeared on LinkedIn. Feel free to connect here.