Do you know where the single biggest dropoff is in your current onboarding experience?

Onboarding is a murky term. Sometimes people mean your first session with a product. Other times it refers to those popups, modals and guided welcome tours you have to endure. Other people think of onboarding as the initial two weeks with a product, usually the free trial territory of the experience.

For the sake of this little post, I would like to define onboarding as everything that precedes someone successfully building a habit around your product.

To keep things clear, I’m going to define ‘building a habit’ as doing a specific action a set number of times withing a specific timeframe.

You’re not pulling these numbers out of a 🎩. You have to sit down with your data lead. Figure out how many times people must perform your core action within an initial timeframe to stand a chance of becoming a long term user.

The only way to do this is to run correlations between how many times people perform an action within different timeframes and how likely they are to continue using your product six months later.

Once you have a placeholder for what people need to do to stand a chance of building a long term habit around your product you have an end point to work with.

The idea is to start at the endpoint and work backwards. Map out all of the touchpoint people must go till you get to when they sign up.

You can measure exactly how many people made it to each touch point. Mapping your onboarding experience in this way lets you highlight the single biggest dropoff in the journey people take to build a habit with your product.

Visualising your onboarding experience as a funnel and measuring where the most significant drop-offs are is a valuable way to find high impact opportunities for growth.


Other questions to ask yourself to improve your retention…

  1. Does your product make a clear promise?
  2. Do you know how often people have the problem you help them solve?
  3. Do you know what your product’s core action is?
  4. How many people continue to use your product six months after they sign up?
  5. Have you segmented your retention curve by the different types of users in your app?

Segment your data by user intent

Let’s say you have a food delivery app. One use case is ordering lunch at work, another common use case is young couples ordering dinner at home, a third one is ordering food for a dinner party.

The time of day and the order size give you enough to bucket most people fall into one or more use case.

The problem is that I threw a dinner party recently and I ordered two of my favourite dishes and we cooked the rest ourselves. There’s no way your app could have known I was throwing a party.

After thinking about this a lot I’ve come to the conclusion that if you can’t tell, it doesn’t matter. If you define the dinner party use case as ordering 8+ items then it’s only relevant when I order 8+ items.

The key here is to define your use cases from the business’ point of view. Then you have to establish unique criteria to measure each use case. The payoff is that you can then segment your behavioural data by user intent.

I’ve found this way of thinking about segmentation helpful because it leads to a productive conversation about relevant product changes.

For example, we now know that people who order lunch at work are the most likely to continue using our app six months later, as compared to our other use cases. How can we make our product better for this use case? When can we start interviewing people in this segment to understand their sticking points? Does personalising the product experience make sense here? Could we run an experiment to see if we can encourage people at work to order lunch as a group?

Compare this to aimlessly segmenting your data by easily available information. Looks like most of our users prefer Chrome…on a laptop…in the US. Interesting.

Segmenting your data by user intent helps tame what can sometimes be an ocean of behavioural information to highlight regions of insight that can lead to productive conversations about relevant product changes.


Other questions to ask yourself to improve your retention…


User Retention

Retention is a measure of how often someone comes back and uses your app after they first sign up.

It matters because no other aspect of product work connects as many important dots.

Retention directly affects how much money you make. If people stick around for long enough to pay for another month, that’s just money in the bank. The longer they stick around the more chances you get to upsell them onto higher plans or other services.

The longer people stay the more likely they are to bring others into the product. I’m not just talking about people inviting their friends, I’m talking about virality as a consequence of natural usage. People signup, they use it to do whatever it was designed to do, if your product works then the thing they make goes out into the world, people see it, they like it, they check out your product, some of them sign up. 

Not all products have this kind of virality baked in, some rely on ads or sales teams to get people in the door. This brings us back to money in the bank. The more people you retain, the more money you make, the more you can spend on ads and sales, and round and round it goes. 

Retention is deeply connected to every aspect of a business and it is one of the most underrated growth channels. It’s also the least understood. 

Improving retention all begins with measurement

The question you want to be able to answer is how many people continue to use your product 6 months after they sign up?

Before you can measure the number of people who use your app you have to define what “using your app” means.

Meaningful usage implies that people are using your product to solve the problem they signed up to solve.

For Airbnb, meaningful usage would mean booking a night. For WhatsApp, it’s probably sending a message. If you do rideshare it’s probably booking a cab, if you deliver food then it’s likely ordering dinner. 

The action you’re are looking for is the thing people do in your product to deal with the problem they signed up to solve.

Once you know what your core action is then you need to figure out your usage interval. Picking the right usage interval comes down to understanding how often the problem naturally occurs.

If Airbnb looked at how many people booked nights every week, it would look like nobody uses their app more than once. People go on holiday once, maybe twice a year. It makes more sense to look at how many people book nights every year. What looked like lots of people leaving, now looks like lots of repeat business.

Once you have a core action and your usage interval you can construct a cohort chart.

The first column in a cohort chart corresponds to a time interval. If your usage frequency is monthly then it groups people by the month they signed up. The row shows you what percentage of people continue to perform the core action each month after they signed up. 

Cohort charts are great for spotting changes from month to month, but they are confusing to look at.

I prefer to average all the columns out and plot them on a retention curve instead.

Question I still want to answer

Have you segmented your retention curve by the different types of users in your app?

If you’ve plotted a retention curve and you know how many people continue to use your product six months after they sign up then the next step is segmentation.

Segmentation means looking at subsets of different types of users. Let’s say, on average 40% of people who signup continue to use your product six months later. If you break it down you might discover that 60% of people who sign up with a business email address continue to use the product. On the other hand, only 20% of people who use a free email account (like Gmail) stick around.

You can use this information to focus on your strengths and cut your losses. “We now require everyone to have a business email to sign up”.

The other way to look at this is to focus on improving the segments that pull your average down. Segmentation reveals where you’re underperforming. It maps out regions of fertile ground for growth.

The easiest way to start segmenting a retention curve is to use the data your tools automatically track. What technology people use, their general location, which landing page they came through, the traffic source they came from, etc.

Segmentation gets more interesting once you start to look at more specific user-generated inputs. The free vs paid email is one example here. The actions people take in their first two weeks is another, the notifications they respond to, the ones they ignore, how often they interact with customer support, the features they use, the one’s they don’t, how often they use the product, their payment tier, etc.

The idea is to segment your curve in as many ways as you can. Each layer of segmentation tells a different story.


Other questions to ask yourself to improve your retention…

  1. Does your product make a clear promise?
  2. Do you know how often people have the problem you help them solve?
  3. Do you know what your product’s core action is?
  4. How many people continue to use your product six months after they sign up?

How many people continue to use your product six months after they sign up?

Before you can begin measuring the number of people who use your app you have to define what “using your app” means. Meaningful usage implies that people are using your product to solve the problem they signed up to deal with.

Meaningful usage for Airbnb would be booking a night. For WhatsApp, it’s sending a message. If your app does rideshare then it’s probably booking a cab. If you deliver food then it’s likely ordering dinner. 

Once you have your core action then you need to figure out your usage interval. Your usage interval is based on how often the problem you solve naturally occurs.

If Airbnb looked at how many people booked nights a week, it might look like nobody uses their app more than once. People go on holiday once, maybe twice a year. It makes more sense to look at how many people book nights a year. What looked like lots of inactive users, now looks like lots of repeat business.

Once you have your core action and your usage interval you can construct a cohort chart.

The first column in a cohort chart corresponds to your usage interval. If your usage interval is monthly then it groups people by the month they signed up. The rows show you what percentage of people continue to perform the core action each month after they signed up. 

Then average out all the columns out and plot them on a curve.

Now you know how to work out the exact percentage of people than continue to use your product 6 months weeks after they sign up.


Other questions to ask yourself to improve your retention…

Do you know what your product’s core action is?

The whole point of a core action is that when people do it they're more likely to continue using your product in the long run.

Your core action is the thing people do in your product to deal with the problem they signed up to solve.

You want to start by shortlisting all the things people can do in your app to address their core motivation for signing up. 

To narrow it down, think about how often people experience the problem they’re dealing with. If the problem is filing your taxes, there’s not a lot you can track on a monthly basis that will tell you if people are successfully using your product to file their taxes this year. A yearly metric makes more sense here.

Airbnb use yearly metrics. People only go on holiday one or twice a year so it makes sense to track how many people book nights each year. 

On the other hand, a chat app like WhatsApp helps people coordinate on a daily basis. Tracking how many people send messages each day gives you a clearer picture of how useful it is.

A good core action will line up with how often the problem naturally occurs.

If you have data to run a correlation then the last step is to correlate your contending actions with long-term usage. The whole point of a core action is that when people do it they’re more likely to continue using your product in the long run.

If you still have options after you’ve run your correlations then always go with the metric that’s easiest to understand. A good metric that’s simple to grasp is better than something more accurate with lots of footnotes.


Do you know how often people have the problem you help them solve?

The natiral frequency of the problem

A product is meant to help someone solve a problem.

Understanding when your problem occurs lays the foundation for how often people are expected to use your product.

If people deal with your problem once a month, sending them emails every day is going to feel like spam. On the other hand, a monthly reminder is will probably be appreciated.

The natural frequency of the problem your product solves is the bar you use to gauge what’s too little and what’s too much. 

Establishing this frequency isn’t an exact science, you’re going to have to ballpark it.

You could always look at how often people use your product and extrapolate from there. The problem with this approach is that people might be using it less than they should.

“Most of my users log in twice a month”. Twice a month, that’s the bar. If your product is designed for a problem people have two or three times a week then you’ve got loads of work to do.

It’s easier to benchmark usage with your best guess of how often your problem naturally occurs.

You can refine this measurement by thinking about other ways people solve the problem. How do people who don’t use your product solve the same problem? How often do they have to deal with the issue?

I’ll close this up by saying that you might have lots of different use cases. Some people use your product to set things up once a month, other people use it to sort stuff two or three times a week.

Pick your primary use case and focus on that for now. Then repeat the process for all your other use cases later.


We There Yet?

Rather than asking if we’ve released the feature, we should be asking if we are getting closer to the outcome.

This means starting the conversation with an outcome, to begin with.

How’re we supposed to build something that drives an outcome if we don’t know what the outcome is? How’re we measuring it? How will we know when we’re done?

Once we have a goal post, we can begin exploring opportunities that will drive the outcome.

If my desired outcome is to improve onboarding, I want to know two things. First, what stops people from activating today? Second, what are my most successful customers doing that everyone else isn’t?

Start with an outcome, explore the opportunity space, now we can talk about solutions.

We must ask if a solution delivers on the opportunity in a way that drives our outcome? If we solve the problem, but it doesn’t improve activation, then we didn’t actually create any value for our business.

I have no idea if we’re going to improve our onboarding, but here are the opportunities that I see, and these are the solutions I’m exploring. Now everyone understands why we’re building this feature and what outcome it’s supposed to drive.


Full credit to @ttorres for this gem on product discovery. She talks about it in this wicked little talk from back in 2016.

Reading to retain customers

I often get asked how I learned about user retention and what resources are available to learn more about retaining customers.

I’ve put together the following reading list to help expand your understanding of how to retain customers. This list is not exhaustive, I’ve limited the recommendations to the most useful ones I’ve found so far.

Amplitude’s playbook on mastering retention

This is the first resource on how to retain customers that I discovered and completely fell in love with. It’s an excellent introduction to the topic. The book is practical. As a result, it shows you how to define your core action and set up your analytics instrumentation so that you can build a retention curve. Moreover, it’s completely free.

Reforge’s retention + engagement deep dive đź’°

This one’s not free. However, it is the single most comprehensive resources on how to retain customers I have ever come across. It’s a phenomenal 6-week program and it costs $3500. The program is fairly hard to get into. You must have a few years of experience in the industry to apply. Above all, you must also be working on a product team when you apply. If you don’t meet these criteria, the course will be an expensive waste of time. If you meet the criteria, there’s nothing else quite like it. I am grateful to have been part of this program.

Udacity’s course on activation and retention strategy đź’°

This program is jam-packed with useful information. It’s an excellent next step if you’ve read the Amplitude playbook and want more.

The elements of user onboardingđź’°

This one is not specifically about how to retain customers. However, a large part of retention comes down to improving your onboarding experience. I have the utmost respect for Samuel’s work and perspective, and his book comes highly recommended. If you want to improve how you think about your onboarding experience look no further. Sign up for my free workshop and you’ll also get a link that gives you 20% off the book.

What is good retention by Lenny Rachitsky

As soon as you start working on retention, you will begin to ask yourself what good retention mean. I will point you to Lenny’s Benchmarks here. I continuously go back to this reference when I want to know what good retention means for products in different industries.

Brian Balfour on how to retain customers

I will leave you with Balfour’s talk on retention at CXL from 2016. I’m sure he has done even better presentations on the topic since. This video is where I first discovered his work and it remains my favourite.

There you have it, my top resources on how to retain customers đź‘Ť

If you know any other resources that you think should be on the list please let me know.

Web form design best practices

web form design

On most sites, web form design is tied to conversions. Form optimization is a high impact activity to pay attention to.

Set clear expectations

Fewer fields are not ALWAYS better

Sometimes asking for less decrease conversions.

Not asking relevant questions can cost you credibility.

Track form completion so you can tell the difference between removing unnecessary stuff and clarifying how to complete important stuff.

Long forms let people self-qualify

Deliberately increasing the number of fields improves quality.

A shorter form means more people will complete it. Only the most motivated make it to the end of a long form.

Breaking forms into multiple steps reduces perceived friction. If your form contains more than 8 fields, try a multi-step form.

Auto-fill where you can

For example, ask for a zip first so that you can derive the address.

Tell people too, “If you add your zip first we will fill in the city and state for you”

Browsers can auto-fill a lot of common form info these days. Use this.

Nobody wants to see your error message

Your top priority is to avoid errors.

Now go fix some forms.