Building a Retention Strategy, Part 1: Identifying your Active Usage Metrics

Today, everyone from the C-suite on down has retention on their minds. With the tech economy slowing down and companies tightening their belts, it’s never been more important to keep your customers happy. 

While it’s generally recognized that retention is a cost-effective strategy, for many companies it is a newer motion. This can be especially true for high-growth companies, who until now have been laser-focused on acquisition. 

To download the worksheet for this post, go here.

If you’re worried about developing an effective retention strategy, this series is for you. In this and future posts, we’ll walk through the key ways to use data to make sure you’re building a smart, scalable retention motion. Most will also include worksheets you can download and use with your team.

Future posts will focus on identifying leading indicators for retention, tying usage metrics to financial metrics, and prioritizing your book of business engagements & interventions. For this one, we’re starting at the beginning: identifying your active usage metrics.

Let’s dig in!

PART 1: Identifying your Active Usage Metrics 

Developing a smart retention motion involves taking a hypothesis-driven, iterative approach.  Identifying your active usage metrics is the first step.

Why? Because before you can start figuring out how to make sure your customers are getting value, you have to make sure: 1) you can tell if and when users are getting value, and 2) you’re able to track this reliably over time. 

Setting this up means discerning the behaviors that you want to see people doing on a regular basis. Doing this provides the foundation for understanding what keeps your high-value customers coming back, and for figuring out what your team needs to do to improve retention for different categories of users.  

The Goal: Figure out which actions, behaviors, and events in your product equate to value for your business.

Once you have your indicators defined, you can tie them to financial metrics, then figure out which metrics are leading indicators for retention. We’ll discuss this in later posts.

But before any of that, we have to define active usage. It’s time to bring your team together and brainstorm—come up with some hypotheses, then gather data on them. 

How to identify active usage metrics: a step-by-step guide  

Let’s do it! All you have to do is set aside a few hours with your team and follow the instructions. We’ve broken the process down into 7 easy steps, along with an accompanying worksheet you can print out. (You can download it here.)

Note also that this process doesn’t have to be a workshop, though it does lend itself to that kind of process. But if it’s hard to get your team together, you can also do these activities async. 

The steps:

1. Brainstorm

Hypothesize: what signifies getting value in your product?  

Put your team in a room and spend at least 10 minutes on this. Your goal is to come up with a list of activities in your product that best capture getting value.

Think about your users' needs, preferences, and pain points. What indicates that users are using your product in a way that benefits them?

At this point you can be general. Later you can figure out what specific events in your product correlated with these activities. 

The specifics will vary according to the nature of your product, but here are some examples to get you thinking:

Go nuts! Use a whiteboard or sticky notes and cover the walls. 

At this point, there are no wrong answers. You can always whittle them down afterwards, but this is the place to get creative. Brainstorm hot, edit cold!

2. Top 5

Arrange your list into a Top 5.

As a team, vote on the top 5. It doesn’t matter so much how - you can run a multi-vote, put stickies on a whiteboard board, take an informal poll, or whatever works for your team. The goal is just to have a list of the main activities you want to drive people towards.

Examples:

1. Create user profile

2. Submit payment information

3. View product tour video

3. Define Behaviors

For each activity, write out the behaviors that define it.

Here’s where it gets specific. For each of those top 5 activities, what specific events in your product represent a user’s (or team’s) completing this activity?

Is there a button that needs to be clicked, a form filled out, or a page viewed for you to know they’ve done the thing you were hoping they would do?

Examples:

4. Figure out frequency and persona

Ask yourself who needs to do what, and how often.

The next step is to go through these events and decide who should be doing them, and how often. Again, this will vary by product.

Which type of user should be taking this action? Usage rates can vary greatly by persona. Executives may only log in occasionally, while you may want your power users to be logging in every day.

When thinking about retention, think about which users are most important. Which users are most likely to see your solution as necessary to their job? It’s their actions you want to focus on.

How frequently should users be doing these activities? This too will vary. Once you’ve figured out who your users are, think about how frequently you’d expect them to take this action. Again, the question is: what frequency best captures getting value from your solution? Or: what frequency is the minimum you’d expect to see to know that your solution is a key part of a user’s workflow?

In Edtech, for example, you might want people to complete a weekly assignment. In a tax product, the frequency might be yearly. In social media, it might be hourly…or even by the minute!

Ok, got your personas and frequencies lined up? GREAT! 

Now you have some strong candidates for active usage metrics. 

5. Analyze your options

Now use your analytics tools to see which are most meaningful.

Ok, time to dig into the data. Fire up your analytics tool and start investigating.

First, take each candidate event. See how often those events take place. Then do some grouping to see how different customer types interact with your solution.  

• What are current customers doing differently than churned customers?

• What do your best customers do? 

• What do the worst ones do?

Look at the rate and frequency of all of these behaviors. Be sure to examine them by time period.

The goal here is to find actionable differences between groups. You want to start figuring out what event(s), at what frequency, correlate with ongoing usage.

Note that you don’t have to find the exact answer now. You’re just trying to isolate quality candidate events. These are your hypotheses about which event, performed by whom, and at what frequency, will most designate getting value from your product. 

It’s a great idea to create an engagement matrix to visualize the results. This is a simple quadrant graph that will let you gauge exactly where users land on the spectrum of interest. 

It shows how many users do an event, and the average number of times those users do it. You can use this type of chart for many purposes: 

• Exploring your most and least-used features

• Measuring feature success across the board at a glance

• Seeing how engagement varies among segments (such as mobile vs. web users)

Here’s a sample engagement matrix:

6. Choose

Decide on ONE metric that is your primary usage metric. 

This will be the metric you use to determine whether people are getting value or not. 

(It doesn’t mean you’ll try to get people to ONLY do this action. Making other parts of the product easier to use may result in more people doing the desired thing.)

Hopefully, the data will reveal this for you. But if not, you may have to decide as a group. It may be a qualitative decision. 

Again, this may not be the final metric, or frequency, you end up with. As with everything, you want to iterate over time. That said …

Once you have it, commit to it. Now you’ve got a North Star!

7. Share

Get a consensus. Ask outside your team.

Trust the data, and your gut—AND get second opinions, too.

Remember: As long as you’re being creative with your ideas, then rigorous and methodical about testing them—including recognizing when hypotheses are wrong and changing directions as needed—you’re doing this right.

If you want to download the worksheet version of this post, you can do that here.

There’s more to come in our retention series!

In our next few posts, we’ll talk about:

• How to identify the leading indicators for retention.

• How to tie your usage metrics and retention indicators to financial metrics.

• How to prioritize your book of business engagements & interventions