The ultimate guide to understanding and improving customer retention

For all of my businesses, I’m focused on one main thing. How do we provide a great product (or service) and keep people coming back again and again.

For some things that means focusing obsessively on the product. For example with Black Hops brewing we have only put one commercial beer out but we have brewed 10 different pilot batches of beer and learned something from every one.

RelatedLessons learnt brewing 10 pilot batches of craft beer

For projects like my book, it was going back and forth literally hundreds of times with the (very patient) editor to get it right, putting out hundreds of related content pieces for free and starting and engaging daily in a community to help people after they buy it.

Product is important for monthly recurring businesses as well. However getting someone to pay for a product is one thing. Keeping them staying on is another.

This post is a collection of ways to keep people on as customers. It’s a general list and not something I’ve specifically tested in any of my businesses. However if you are looking to improve customer retention, you should get some ideas here.

1. Start measuring churn

To increase retention you first need to understand and measure how many customers are leaving. There are 3 main things I’ve picked up on this topic.

Basic overall churn

A simple measure for overall churn is:

No. of people who have left this month / Number of customers at the start of the month.

So if you have 50 customers at the start of January and 5 customers left in January, your churn rate would be 10%.


This is a simple and somewhat useful overall churn metric, that comes with some major issues. One that stands out is churn will often be extremely high in the first month of service.


Say you start a membership site and someone signs up knowing full well you offer a money back guarantee. They log in, decide it’s not for them and then they cancel.

This is very common and it’s not unusual to see first month churn rates as high as 30% for certain types of businesses. More often than not, this is not a customer retention problem, it’s one of many other possible problems. It could be simply the copy on the site overstates the value and people sign up expecting more than you can offer. Perhaps they never really intended on sticking around and have only signed up to get some key resources.

Of course all of these things are issues you want to deal with, but they aren’t retention issues. Yet when you look at an overall churn statistic, if your 1st month churn is 30% your overall churn will be bad also. You may or may not have a retention problem.

Churn Grid

I chatted to Rob Walling about this at length at a conference in Bangkok in October 2014. He showed me his Churn Grid for his software apps which looks like a better way to look at churn.

The churn grid shows churn rates on a monthly basis. For instance for month 1 it might show that you have a 30% churn rate. But then it will show how many people churn after they have been a customer for 2 months and 3 months and so on. To look for retention issues, you need to be looking at the latter months.

If you are in a situation where your 1st month churn is very high and your latter month churn is low, and you are growing very quickly, then your business isn’t really at risk. Next month for example you might have 50 people sign up and 25 of them leave. The churn rate will be 50% for that month and the overall churn rate will be very high as well. But if your current customers stay, you will be no worse off.

This is a common situation for high growth companies. The growth itself is over inflating the churn rate.


Cohorts may be the best actionable way of going about measuring and managing churn. A cohort is simply a group of people who became a customer under certain conditions (time being the easiest one to understand).

Let’s say in February you decide to test whether or not a signup fee will increase overall retention. The assumption is that if you make people pay up front, they will feel like they have ‘invested’ and they will be less likely to cancel later. You might even scare off the customers who are more likely to churn quickly. So in February you add a $200 sign up fee and create a ‘cohort’ of people who have joined your service under that deal.

3 months down the track you can look at how this cohort performed against your normal customers, or another cohort who you offer a better deal to. You can look at whether or not the people who paid the sign up fee were more likely to stay for a longer period and make an informed decision around whether to stick with the sign up fee.

You can do cohorts on a time basis without any complicated software (as described above). You you can use analytics software like Kiss Metrics to split groups up by offering different flows to different people on your site right now.

RelatedLearning More About That Other Half: The Case for Cohort Analysis and Multi-Touch Attribution Analysis

So essentially with this approach, you make a bunch of assumption about what might keep customers around longer and you then test those assumptions with different groups. The conditions presented to the group that performs the best, are the conditions you run with (sign up might be one example).

What is an acceptable churn rate

One of the hardest things about churn is it’s extremely difficult to work out if you actually have a retention problem. Every business is totally different. If you consider a hosting company for example. They most likely have a very high 1st month churn, because people will sign up and find it too hard to move their site across and give up. However long term, if you have your site hosted somewhere, and they are doing a good job you probably won’t leave. These sorts of businesses can pay a lot of money up front to acquire customers. CRM systems and other apps where you invest heavily with your own data, are in the same boat.

Memberships and service businesses don’t have much of a barrier for people to leave. For those types of companies, long term retention is going to be a challenge.

If you look at the popular content out there on churn, you’d believe that something like 5% per year is an acceptable rate of churn. This is less than half a percent per month and from my experience, most people I know would kill for this level of churn (Even software companies).

Here is the monthly churn rate for Stripe analytics dashboard, Baremetrics. The publish their metrics live here. It’s the same tool used by Buffer to publish their metrics live here.


Unfortunately I can’t come up with better guidelines around what churn is acceptable except to say this. Any customer that leaves is one less person that will pay you, one less person that will tell others about you and more importantly one more person who will use a competitor and possibly tell other people that they left you and chose someone else. It will also reduce your overall customer lifetime value which lowers the amount of money you can afford to pay to acquire a customer, putting you at an immediate disadvantage in relation to your competitors.

So unless you have zero churn, it’s worth paying attention to.

If you are running a high growth company, as long as you are signing up more clients than those leaving then you are ok for now. Longer term though, if you have a high churn rate for existing, activated customers then it will come back to bite you.

9 ways to improve customer retention

For the next part of the post, I’ll delve into the ideas on how to improve customer retention.

1. Choose a service with a bad alternative

If the alternative of someone using your product or service is not a good one, then they will most likely stay. One example to think about is insurance. People can always ditch their insurance company and take the risk themselves. However despite rarely using insurance, most people don’t take that option. The alternative is not pretty if you find yourself in a position where you need it.

Our course competition is a factor here as well. If there are plenty of good alternative insurance companies to choose from, then more people will move.

If your business is in an environment where people really don’t like the alternative of not dealing with you, then it’s a good start. One example here might be tackling a huge problem area. Uber for example solved all kinds of problems with their service. I don’t want to deal with Taxi drivers not showing up, smelly cars, being ripped off, unsafe driving etc. Uber happens to be cheaper than Taxi’s, but I would pay 3X a taxi fare if the only alternative is a Taxi, particularly if that alternative meant waiting around in an unsafe area for hours.

If you are at the idea stage in your business, this stuff is worth thinking about. Keeping in mind that typically those businesses where no good alternatives exist, are generally big ideas that are not easy to start. Peter Thiel refers to such businesses as monopolies. They aren’t the kinds of businesses you start in a weekend.

RelatedBook Review: Zero to One by Peter Thiel and Black Masters

2. Choose a service with naturally ongoing work

If your product or service is more of a one off need and you offer it as recurring, then you will struggle. Think Dollar Shave Club. If men’s beards didn’t continue growing or blades never went blunt, that recurring business would cease to exist.

It’s not uncommon in productized services for people to leave because you solve their problem too well. They come with a pre-existing issue, you solve it and they leave. Say you offer marketing automation services and part of the deal is setting up Infusionsoft and all of the nurture sequences that come along with it. That work will get you through the first few months but how much work is required in months 4, 5 and 6 or years 1 and 2.

I see a lot of services businesses that include more ‘one off’ type jobs in their description and my guess is that leads to a high level of churn. SEO services is another example. If they are mainly doing on-site work, they are going to run out of things to do after the first month. It might take the customer a while to notice, but they won’t stick around for long once they realize you aren’t doing much.

3. Find out why people are leaving

Jason Cohen says people churn because of 3 things.

  1. Utility – By this he means if it’s not useful to them on an ongoing basis then they won’t stick around.
  2. Service quality – If they have a bad experience, they will go elsewhere.
  3. Pricing – I the product or service becomes too expensive or they can no longer see value, they will go elsewhere.

Utility most likely happens at the idea level. If your service isn’t inherently useful on an ongoing basis, then it’s going to be hard to combat this one.

Pricing is only somewhat under your control. If you have excess margin, you may have some leeway to reduce pricing. Perhaps through internal systems, you can become more efficient to increase margin and reduce prices. However you are still at the mercy of your competitors. If there is an alternative that offers the same thing at half the price, it’s going to be tough to do too much about that.

Service quality is the most actionable thing to look at here. It’s easy to measure and given enough resources, it’s possible to fix. It also impacts on the others. For example if your service quality is much better than competitors, people may pay more to use you.

Here are 3 simple ways to measure service quality:

1. Customer happiness scores

Most help desk systems have a feature now where customers can rate how happy they are. A very simple happiness metric is the % of happy customers minus the % of unhappy ones. For example if you have 88% happy customers and 7% unhappy customers, your score is 81. Improve this score and you will improve churn (assuming that it’s the service quality that is causing it).

2. Net promotor score

The Net Promotor Score (NPS) is a well accepted survey that you can send to your customers. It asks the simple question:

How likely are you to promote our service

From this you can calculate an overall score and track that over time as a metric for your service quality. What I like about the Net Promotor Score is it will factor in things other than just how good your service is. For example if you ask me today how likely I would be to promote Uber, I would say score it at 10.

That has a lot to do with the current competitive landscape. If there was a competitor who came along who were better or if we had self driving cars that were a tenth of the price, I’d be very unlikely to recommend Uber. If you just look at customer happiness scores, you won’t be seeing the whole picture.

RelatedLessons Learned Sending a Net Promoter Survey

3. Ask customers why they left

When someone leaves your service, ask them why they left. If you have software you can build this into the exit process. If not you can ask them yourself or automate a sequence that emails them when they leave. Often something very simple works best here.

‘Did we do something wrong’ is likely to get a response. It’s very direct and a bit leading so people may feel bad that you think they did something wrong, so they reply with the actual reason why they left. I believe this is an idea from Jay Abraham.

A common reply might be ‘No! You guys are great, but……’ and whatever comes after the ‘but’ is your reason.

4. Improve service quality

Regardless of what is causing people to leave and the fundamentals of your business, an improvement in service quality is likely to increase customer retention.

Not only that, improving the quality of the service will have a meaningful impact on growth. If people become happier they will be more likely to refer you instead of a competitor.

In terms of actionable ways to improve quality, these should come from the data collected above. When it comes to services, improving communication, responsiveness and overall accuracy are 3 of the more obvious ways to improve service quality.

5. Increase plan length

A simple way to keep customers around for longer is to increase the minimum plan length.  Here are a few ways to do it.

1. Lifetime plans

John Dumas runs a podcasting community called Podcasters Paradise. While a lot of other people launched monthly subscription communities and battled hard against churn, John didn’t have that problem. He has no churn. Everyone is a customer for life.

It seems like a hyper-aggressive way to combat churn but there is no doubt in my mind that John’s lifetime value would be higher than someone offering the same thing for 1 / 12 of the amount and charging monthly.

It introduces new issues of course and there is still an issue with communities with community members becoming inactive, however from a financial point of view he has his money up front.

It has another benefit as well with communities in that community members leaving can kill a community. I’ve had it happen to me before. A bunch of influential members leave, or it is obvious that people join and then quickly leave. Threads start popping up about ‘where is x’ etc and the whole vibe changes.

2. Annual plans

Not everyone wants to give away lifetime access and doing so can get you into hot water.

RelatedA sustainable business – WooThemes moving away from lifetime support

The accepted approach with annual plans seems to be to offer a significant discount (2 months free or more) to people who sign up annually. WP Engine do this during the checkout process. Leadpages offer a 30% discount to people who sign up annually.

A more aggressive option is to force annual plans. At various times I’ve seen companies including KissMetrics and Hubspot do this. I’m not sure if they are still doing it, and quite often companies that do it don’t make it 100% clear that they are doing it. For example they might say $99 / month and not mention that you have to pay for a whole year unless you delve into their terms and conditions.

I think you have to be careful here, but as long as you are clear that people have to sign up annually, then this could be worth considering.

I’m sure this would also significantly reduce signups to the service as a lot of people won’t want to sign up for a whole year. However you may decide that the cost of onboarding people who just leave after a few months justifies the loss in their business.

3. Quarterly invoicing

A less aggressive, simple option is to invoice quarterly instead of monthly. This at least means you get 3 months worth of revenue which might be enough to make that customer profitable. This might be a good balance in a lot of cases, however it’s obviously not going to keep customers around for as long as ones who have to pay for a whole year.

4.  Lock ins

Lock ins are another way to increase the length of your plans. With this strategy you let people pay monthly bu you force them to stay for a certain amount of time. It’s a bit of an old school mentality and reminds me of Telco’s who lock you in for 24 months just to use their phone service. It doesn’t exactly breed loyalty and trust and usually pisses people off in the long run. That is particularly true if they know they can go elsewhere and they want to but they are locked into your service.

Still, there’s no denying it at least works at keeping people around for that period (long term brand impacts are less clear).

Hubspot may do this I’m not sure. When I fill their pricing form in the price comes out at $33,000 so maybe that’s what they are doing who knows? My guess is they don’t let you sign up just for one month.

6. Pricing or price structure changes

There is a complex relationship between price and customer retention. On one hand having a higher price can deter bottom feeders from signing up in the first place. Those people may be the most likely people to leave, thereby reducing your overall retention rate. On the other hand, as I mentioned above, people can leave because the price is too high.

Price is something that is fairly easy to test, so it’s worth at least trying to figure out the complex relationship between price and your customer retention. A few easy things to try:

  1. Increase your prices and see if you get a better quality of overall customer.
  2. Reduce your prices (if margin permits) and see if your service falls into the ‘no brainer’ category and people stick around.
  3. Charge an up front fee. This might reduce in less signups but people might feel invested to stick around longer. Some companies charge huge up front fees (Infusionsoft were charging $2,000 when I signed up, I haven’t forgotten about it!).

7. Actively boost utility in your service

Since we know utility is one of the 3 reasons for churn, it makes sense to make steps towards making it more useful. As I mentioned above, some of this comes back to the idea level and also some of it isn’t in your control (there may be competitors that are more useful).

There are a few simple things you can do in a services business to be more useful:

Proactively reach out

Every week or month, reach out to your customers manually or via an automated sequence. Ask them if they need help generally or ask them if they need help with something specific. You can automate this kind of thing with automation tools like Infusionsoft if you want to go down the automatic route.

Proactively suggest ideas

Every month you can specifically look at how the customer is using the service and make a specific suggestion for them to do. For example if you are a monthly SEO service, you could look at their recent content and make some suggestions about how they could better optimize their site.

Reactivate inactive customers

If customers haven’t engaged with the service in a while, you could make a special effort to re-activate them. For example if you run a marketing automation service, building sequences and someone hasn’t used the service in 30 days, you could look at their recent content and suggest a sequence to offer as an opt in bribe on their site. This is different to the proactive suggestions because it’s a white glove service, specifically for inactive customers. These customer may be more likely to leave (however this isn’t guaranteed either).

Doing set monthly jobs

Of course if customers have signed up for a service where they get certain jobs performed each month naturally, then simply doing those jobs might be enough. In the SEO example perhaps it’s a monthly job for every customer, for you to analyze all of their blog posts for the month and re-do the meta tags and titles to increase the click through from Google.

Lifecycle emails

Lifecycle emails can also be used to boost utility. Simply sending out a suggestion each week may provide enough value to customers to encourage them to stick around. This needs to be tested though, because if they find that the emails aren’t useful enough, it might just be a reminder that the service isn’t useful for them.

8. Better qualifying

One way to stop people leaving is to not let them sign up in the first place. Again it’s an aggressive way to think about it but if the customer costs more than they are worth, then you probably don’t want them signing up (unless you have another strategy for word of mouth or loss leaders etc).

There are 2 ways I’ve seen this done well.

1. Tempt customers with with a different offer

If you offer a monthly service, you might find people sign up for it who don’t have an ongoing need. They still sign up despite that, because they want to use you and you don’t offer them a reasonable alternative. You might find doing some one off fixed price service offerings is what is needed to get them away from the recurring plan. They still have their problem solved, they still tell people about the great service and they leave happy. The challenge here is to not make it so tempting that no one signs up for the recurring service, and you build a business based off one off projects and not repeating income.

2. Have an aggressive new customer process

I’m a member of a forum where there is a whole bunch of requirements for getting in. I know people who have tried and failed 3 times to get in. You not only need to meet certain criteria but you need to be referred by an existing member. I am willing to bet that the customer retention rate of that forum is much higher than a typical forum. Of course they wouldn’t get as many people signing up at the front end but longer term retention would be much better.

For something like a community where customers leaving is so public, this is a great strategy.

9. Failed charges

The final way you can keep customers around is by aggressively targeting failed charges. Quite often people leave because their card expired and that either prompted them to decide to leave or they just didn’t feel the need to renew.

There are a few ways you can address this.

  1. Allow customers to enter multiple cards and fall back to the second card. Uber does this well.
  2. Use a service that will still work if the customer changes their card (Stripe or PayPal).
  3. Have a sequence that emails people before card expiry.
  4. Contact people when their cards expire and on an ongoing basis after that until they either update it or decide not to (within reason).

I’ve also seen services like ChurnBuster that manage this for you.

What have you done to boost customer retention?

I hope this post has been useful. Let me know in the comments what you think and what ideas you have for boosting customer retention.


Follow me