Getting 200% More Actionable Feedback from Customers that Cancel

Getting useful cancellation feedback from customers is tough.

The problem is that once people have canceled, they’re no longer engaged and will rarely spend the time to give you feedback. One of the best things I’ve ever done to combat this with Bidsketch was to add a mandatory freeform text field that says:

Please help us improve by stating the reason (Required)

Sure, some people enter blank spaces or random characters, but the majority make an effort to say something semi-useful.  A small percentage will write truly useful comments that can lead to some sort of action.

To analyze the data I import the comments into a Google spreadsheet and add a few columns to help me segment them.

The columns:

  • Days Active
  • Plan
  • Cancellation Category

Click below to see what it looks like:

Cancel Spreadsheet

Days Active helps segment by trial users, new customers, and longtime customers. Hugely valuable.

Not all cancellation comments carry the same weight. That’s where the Plan column comes in. This tells me whether I’m dealing with high value customers or not.

Cancellation Category is how I make sense out of all this freeform text.

While this system works, the majority of comments aren’t actionable.

Here’s what the amount of actionable comments looks like:

Cancel Pie Chart #1

Only 27% of comments are actionable in some way. These comments help give direction, but only a few actually go deep into exactly what went wrong for that customer.

Here’s an example:

“Took too much work to set up.”

It helps to know that onboarding could be improved, but what part? Getting integrations set up? Importing clients? Or maybe importing proposal data?

You can see where I’m going here.

I still consider these comments actionable as they help ask the right followup questions (and when conducting usability testing).

Except that I’ve been going on the assumption that the 27% of actionable comments are representative of the 73% that aren’t. Having spent the last couple of months following up on cancellation comments, I found that this isn’t a safe assumption.

So a few weeks ago I set out to increase the amount of actionable comments by changing my cancellation form…

Getting Even More Useful Cancellation Feedback

Over the last few months one of the top reasons for canceling has become:

“Not using enough.”

Interviewing customers I found that they’re typically not using it for a couple of reasons:

  1. Never got set up (during trial or never tried)
  2. Out of business/slow business

The problem with this approach is that only a small percentage of customers can be reached this way. I’m surely missing a lot of useful/actionable data.

So I changed my form to use both mandatory cancellation options, and a freeform text field.

The key is taking the top cancellation categories and using those as options:

Cancel Form #1

And to get even more insight into a specific category, I take people that click on “Not using it enough”, to a second set of mandatory options:

Cancel Form Flow

In the short time that I’ve implemented this, I’ve seen a huge improvement in the amount of actionable comments.

Actionable cancellation feedback has increased by 200%.

It went form this:

Cancel Pie Chart #1

 

To this:

 

Cancel Pie #2

 

Again, the key is using existing cancellation comments to create options used in the form.

Looking at the early data, I see that I wasn’t accurately weighing a couple of cancellation categories. While what I have now is working much better, I’m sure I’ll continue to evolve the process to gain even more valuable feedback.

Of course, something is better than nothing. So if you’re not doing something to continuously get real feedback from customers that cancel, you’re missing out on a huge opportunity to improve your product and reduce churn.


4 Comments on “Getting 200% More Actionable Feedback from Customers that Cancel”

  1. […] first thing we should do is find out why they’re leaving and focus on any actionable feedback. That much goes without saying. But let’s say we’ve done that and now are left with […]

  2. Thanks for this insight, this actually helps in the current project I am working on.

    And now, to sound less spammy, one thought that was bugging me and actually made me to write this comment. It all comes down to wording and might be just a matter of my personal opinion but I guess it might be worth sharing anyway.

    “Please help us improve by stating the reason (Required)” – my initial knee-jerk reaction? F* You sir, asking for help and making it required?! Completely outrageous! And then adding another required field on top of that?! I demand last 30 seconds of my life back now or heads will roll!

    In all seriousness, I see that one of the options there is “business is slow”. Answer like that is possibly the best out of worst case scenarios because it means the reason they cancel is actually not the product you provide but things not working out on their end. They might be a returning customer after all! Getting them pissed at this critical moment is probably not a way to go.

    One last though – it might be a good idea (perhaps this is something you’ve got implemented already) to add “suspend” option for clients with “slow business”. Disable functionality of their account (and stop effectively paying for it) , but let them unfreeze once they are ready to go on.

    Cheers, hope it helps in some way!

    • Ruben says:

      >> They might be a returning customer after all! Getting them pissed at this critical moment is probably not a way to go.

      The thing is, it’s not a problem for 99% of customers. Most people don’t mind having that be required, and are great in the comments. It’s important enough to the business that sometimes you need to be more aggressive in getting valuable feedback like this.

  3. […] We’ll all get much more feedback from our customers that way. Bidsketch gets200% more actionable feedback from cancelling customers by doing something similar: http://www.extendslogic.com/business/why-people-cancel/ […]


Leave a Reply