Before you plan your product roadmap

Main illustration: Pam Wishbow

Are all your users using all the features in your product? Of course they’re not. Let’s talk about that.

When planning your roadmap, and where the team spend their time, it’s useful to ask: “How many people are actually using each of our product’s features?” This is product management 101 and will probably take you a few minutes of SQL, or a couple of seconds of Intercom. A simple way to visualize this is to plot out all your features on two axes: how many people use a feature, and how often. You’ll most likely see trends like this one.
Feature usage
The core of your product is buried in the top right, because that’s what people are actually using your product for. Note: exclude administrative features like account creation, password reset, etc. in this exercise. They’re not relevant here; they’re simply a cost of having a product in the first place.

If you have features in the top left it’s either a sign of a badly adopted feature or they are your enterprise plan features, in which case you’ve thought about this and it’s ok. But you should probably evaluate whether all of your enterprise customers are actually using it.

An even simpler way to think about it is this: What percentage of your customers or users have adopted each feature? You can do this with simple bar charts. Shown here is the dream product, this is the one we all think we’re designing, when we have Photoshop open. Everyone is going to use & love everything, right…
Dream app feature usage
As every product manager knows, the messy reality of a product looks a lot more like this.
Realistic feature usage
How do you get here? You built a product that had a solid messaging, files and a document editing features. These key features were highlighted on your marketing site, documented clearly with great screenshots, included in your product tour, and every new sign-up loved them.

But success can be a lousy teacher. You believed you could do no wrong and that you could build lots more.

So you added a chat room, but it didn’t go so well, you kinda botched the launch of it. Then you added a calendar and that went worse, no one created more than one event, and it’s still not even mentioned on your marketing site. Now you’ve built this time tracking features that’s kind of popular, but only with a certain type of user. This is your product, you need to fix this.

A side note on disruption

Product feature breakdown
If you’re looking at this sort of feature breakdown, what that tells you is that you have an excellent product for one precise workflow, but you’ve added all these other pieces no one uses. I often think of Hangouts inside Google+ as an example here, or primitive document editing (i.e. BUI) inside Microsoft Word.

In these cases you are vulnerable to disruption. Someone can build a simple product, focussing on that one key feature that’s superior in just one way (cheaper, faster, collaborative, easier to use, mobile etc), and you’ll struggle to compete.

What do I do with my feature audit?

For any given feature with limited adoption, you have 4 choices:

  • Kill it: Admit defeat, and start to remove it from your product
  • Increase the adoption rate: Get more people to use it
  • Increase the frequency: Get people to use it more often
  • Deliberately improve it: Make it quantifiably better for those who use it

Roughly you can visualize it like this:
Types of Improvement

Understand why a feature isn’t being used

Understand why a feature isn’t being used
To get more people using it, rank and resolve the issues that are stopping them from using it. This is where the five whys technique is genuinely useful. You might have a situation around users not using your reports feature. Why? They don’t see the value. Why? They can’t show it to their boss. Why? They can’t get it into a suitable good format. Why? Because our export tools aren’t good enough? Why? Because our API doesn’t produce good data. If you ask why enough times, eventually you’ll work it out and get to the root cause.

Diagram why no one is using your feature
Don’t just talk to one customer because invariably things are more complicated than that. You’ll find these blocking patterns over and over again. You can then resolve the key issues, the fundamental things that are actually blocking people from using all of your product.

Note: If you’re wondering how to talk to specific users who don’t use a feature, Intercom helps :)

This is fish or cut bait. Get rid of it or get more people using it.

Increasing the frequency

Frequency improvements are different. It’s about getting more usage from existing users. You need to identify the changes that will get people using a feature more often.

LinkedIn endorsements do this quite well. They added in these one-click multi-directional endorsements where you can easily/accidentally endorse four of your friends for skills they don’t have, and accidentally connect to four more people as a result, which in turn creates more endorsements, more logins, and more single click broadcasts.

What they’re following is a pattern explained by Nir Eyal, author of Hooked, who says habits are formed from a repeated pattern with four key elements…
The Hook canvas

  1. Trigger: the reason the user goes to the product (e.g. you received an email to say a contact had endorsed you for a skill.)
  2. Action: they take in anticipation of the reward (e.g. scroll, search, browse, etc)
  3. Reward: the user gets from taking their action (e.g. seeing a beautiful Pinterest board)
  4. Investment: the user makes which will plant the seed for more triggers (e.g. subscribe, pin, like, connect, etc.)

As an example, Nir points to Pinterest’s hook as the following:
The Pinterest hook


Let’s face it, feature audits sound boring. Maybe rebranding them as the Lean Feature Analysis Hack™ will make them seem more hip. Regardless, a feature audit is a powerful tool. Every development team on earth has limited resources, knowing what users actually do in your product lets you focus your work on areas where it has impact. Similarly iterating on a feature knowing whether you’re trying to increase adoption or frequency lets you easily measure the results, and avoid valueless pixel pushing.

For more mature companies it can serve as an early warning system to ensure you’re not polishing features no one wants while leaving yourself ripe for disruption by a more focused competitor.

Want to read more of our product best practices? Download our free book, Intercom on Product Management. It’s recommended by folks like Ryan Singer, Hunter Walk, and Dharmesh Shah.