Reducing churn with a single email

I really think it’s time for us at MarketGoo to start sharing some of the learnings we’ve had over the last months. There are many other teams creating super useful content and some of it has helped us greatly to build our project as it is right now. We believe it’s important to invest in giving to back to the community so we’ll try to walk you through our path. Note that we are not experts and here we’ll be humbly talking about stuff that has worked (or not) so maybe it can be useful for you too.

Now, let’s start from the very beginning. As a SaaS company with an SEO product aimed to help SMBs, we have countless challenges on the table. It’s easy to get overwhelmed and lose the north star. But the product team has always had clear the one metric that matters, which is customer churn (or retention). This measures the rate of your subscribers that cancel (or renew) their relationship with you. When I joined MarketGoo three years ago I knew almost nothing about all of this, but Wences helped me understand this was going to be our primarily goal (and the cause for some sleeplessness too xd).

One of our firsts strategic sessions during our retreat in Menorca

We’ve been working hard from different levels to act on reducing churn. However, there was this tiny little workflow we set up that I think was an eye-opener to shape our vision towards churn, user experience, analytics, etc. Reading about other companies I got to discover Groove and its blog where CEO Alex Turnbull was providing much value with articles like How We Grew Our Customer Exit Survey Responses by 785%. I decided to replicate his ‘Open-ended Question’ idea because I felt we were blinded to the real problems and were pushing for solutions from intuition.

The goal here is to start getting useful information about churned customer’s opinion with just a single email, when the customer cancels. Easy as pie. We were measuring quantitatively churn rate (i.e. October churn rate is 10%) but it was difficult to get insights on the whys, the problems. Qualitative data, the one you get by speaking to a single customer, is key here. Qualitative and quantitative data should go hand in hand.

The email

What we did was to create a user auto message in Intercom that is sent to our direct channel paying customers (subscribers) once the ‘cancel-subscription’ event is triggered. And the email goes like this:

As you can see it is pretty straight forward but we’ve been polishing it over time. I’d like to highlight the following:

  • Make it easy to be set up. We used Intercom but I’m sure there are plenty of tools that’ll do the work. Just make sure you don’t rely on tech team so you have flexibility to update content, change settings and measure results.
  • Make it personal. We found out that customers are more predisponed to answer back when there is ownership in the email. For a customer it’s not the same to have this from a support agent (because it’s their job) than from the CEO or founder, taking time to speak personally. In our case I used myself because I wanted to test this without involving our founders and it worked fine.
  • Obviously, answer all of them. You won’t find magic answers that will help you solve your problems but what’s important here is to start having conversations about your product and its experience. What’s making people cancel. Take time, ask the right questions, let people speak and listen. I always try to set up a quick call. If it gets unmanageable, think of a solution. Have it running for just some weeks.
  • Follow back. How cool is it to come back to a customer that canceled because something was not working or there was a missing feature, and now it is solved? Quite motivating and surprising actually, to see the results whether it is the customer reactivating an account or just thanking you for taking the effort. I have to admit we have not mastered yet this workflow but hopefully we’ll get there soon.
  • Measure its impact in the short term. Act on analytics for the email itself. Is it getting answers? What’s the open rate? Think of this as an experiment that needs to be polished too.

The performance

I’m not an expert in email marketing but for a cancel subscription email I’m happy to have been able to take the open rate to 47%. No click through rate whatsoever of course, as there is no CTA placed. But an interesting 13% reply rate with customers actively telling us reasons why they decided to cancel their subscription and opening opportunity for a deeper conversation. (The 175 sent is not the total number, just the last version that is in place from A/B testing).

Would you like to see some examples? I bet you do! 🙂

Well, from these examples you can see some are quite actionable and others need further understanding. However, there is a high risk of getting crazy with this opinions and start taking decisions from them. That’s something you should completely avoid. These are real but isolated opinions. We shouldn’t be planning yet on this feedback.

The next steps

Unless you have a big team and enough resources it can’t get difficult to get completely accurate and rigorous in turning this data into projects aimed to reduce churn.

What I think you can do here in a lean way is first, understand that qualitative data can be dangerous if you don’t understand that single opinions are just that, opinions. For example: Is it a priority for us to invest time enabling Pay-pal billing within the sign-up flow? We believe it is not, at least right now. Our hypothesis is that there are other priorities that could impact the churn metric for the better.

So here I’d like to quote Dan Olsen’s “going quan on qual” from The Lean Product Playbook. This means working your customers answers a bit more, taking the data into a spreadsheet, see patterns, research within your product what’s happening there, talk to the team and see what happens to the rest of the customer base that is not complaining over email. You’ll get wiser and will start drafting hypothesis on the problems. Problems that can be worked out to have a solution to be tested in front of the customer. You got it!

Eventually we have discovered 4 main reasons that drive cancelation. Two of them are being worked now, the third is a bit out of our scope and the fourth is related to any company doing business with SMBs, which is the death rate. There is a % of them that eventually close the business and will no longer need you so it’s difficult to impact here.

Conclusion

Obviously this email is not the Holy Grail that will solve the churn problems all by itself. However, I firmly believe it is a great starting point for you if you haven’t yet get deeper into this kind of challenges or you are just in the first steps of doing so.

Over the months we have taken so many useful opinions that has turned into hypothesis and tests, some have worked and some haven’t but in the long term, it is helping us to assure the path, pivot or recalibrate the product vision when it has been needed.

I see this tiny project maybe as a tip of the iceberg to any starting product or success team.

To us it has meant to reduce considerably our churn rate in the long run:

All time MarketGoo direct churn

I also published a video in my Youtube channel talking about all of this: Reducing churn with a single email. (It’s in Spanish, sorry).

Hope you liked it and can be useful for you. We’ll keep sharing learnings so feel free to tell me what you think or ask anything you need.

To see what this is all about take a look to our recent post talking about why culture is more important than a company.

Thanks!