8 Steps to Building Features Your Users Actually Want

| January 29, 2015
Sign up to get weekly resources, and receive your FREE bonus eBook.
Thank you!

Get ready for some great content coming to your inbox from the team at UserTesting!

Today’s guest post was written by Craig Morrison, founder of UsabilityHour.com. Enjoy!

features header


That’s a quote from Lord of the Rings, in case you’ve been living under a rock for the last… 60 years.

This is the quote that echoes through my head every time someone puts an idea for a new feature on the table.

I would like to scream it out loud, directly at that little post-it note with the feature written on it, but, you know, that would be crazy.

Alright fine, I’ve done it before.

Features are a doubled-edged sword. Releasing a great one could be the defining moment that takes your product to the next level.

Explosive growth, overjoyed users, investor money raining down from the sky… (maybe)

But more often than not, that new feature just ends up as another item in your ever-expanding navigation bar. One that’s distracting your users away from the main reason they’re there.

As the list of features you add to your product grows, so does the impact they have on your UX, on your budget, and on the effectiveness of your team.

You need to be certain that new features will help your product more than they hurt it.

[Tweet this.]

For example, if you’ve promised your users a faster way to manage their Twitter feed, and you decide to add in Facebook support, you’ve changed the value proposition of your product.

Your users signed up for a tool focused soley on Twitter, and suddenly they’re finding features for Facebook mixed in. This can be lead to confusion and frustration when they have problems locating the actual features they signed up for.

Going a step further, these new Facebook features now need attention from your team. Bugs need to be fixed, support cases need to be dealt with, functionality needs to improve.

This means you’re spending money, and your team’s time, on this additional Facebook support.

But what if no one even wanted it?

What if Facebook support is getting in your users’ way, costing you money, and taking focus away from improving the Twitter functionality?

What if it was just the result of a single user saying… “Can I use it for Facebook too?!”

Sounds like someone should have screamed “YOU SHALL NOT PASS” at that post-it note.

Car covered in Post It notes

“Frank’s wife told him to stop bringing his work home with him, but he was all out of whiteboard space.”

This is why feature requests need to go through a strict process that ensures they help your product more than they hurt it.

I’ve seen firsthand what the impact of a runaway train of feature development looks like. (Hint: it looks like an empty bank account.)

So I’ve put together 8 detailed steps for you to follow to ensure that each feature request meets the following criteria:

  1. It solves a problem your users actually have

  2. It’s usable

  3. It isn’t creating clutter

Note: Strapped for time? I’ve put together a free checklist that summarizes “8 Steps to Building Features Users Want” to save you time, and give you something to easily refer back to later on. You can download it at the end of this article.

These 8 steps are entirely based on Ash Maruya’s feature building methodology from Running Lean.

If you find these steps useful, I highly recommend grabbing a copy of Ash’s book.

1. Set Your Default Response to “No”

As your product grows and you start to attract more users, the floodgates open up on feature requests.

They will come from everywhere. Your users, your team, and especially your own brain.

Why? Because thinking up features is AWESOME. It’s the fun part. You get to visualize how users will benefit, and how it’s going to be a game changer.

The problem is, everyone’s idea for a new feature is a game changer. It’s not even possible to change the game that many times!

It’s important to set your default “response” to NO.

I like to think of feature requests as worthless until proven valuable. The moment they are born, they are put on trial for the attempted murder of your product.

Think of feature requests as worthless until proven valuable.

[Tweet this.]

What follows is a series of trials which attempt to prove them valuable. Fail to prove its value, and it pays the ultimate price.

Adopting a mindset that new features are bad will limit your excitement and enthusiasm, preventing you from charging forward and implementing them without the proper research and validation.

A new feature standing trial

“Mr. Post-it… did you, or did you NOT, overstate your value to the product?”

2. Identify the Problem the Feature Solves

I love this step, because this is a “kill step”, meaning that if the demand for the feature can’t be proven, this is the first point at which we can wipe it off the list.

In order to be valuable to your users, the new feature needs to solve a problem for them. A problem worth solving.

If it doesn’t, they won’t use it, and then it becomes clutter. Clutter takes away from their experience with your product. Clutter = bad.

A new feature needs to solve a problem that’s actually worth solving.

[Tweet this.]

This first step might be the easiest, yet most important. It’s as simple as talking about it and trying to justify why it should be included.

Find the source of the feature request. Was it a user? Was it a team member? Either way, sit down with the user, or your entire team, and have a discussion around the reasons for building it.

In Running Lean, Ash Maruya suggests, “Try to talk the customer out of wanting the feature. Have them sell you on why it should be added.”

Determine if this problem is actually worth solving for your users. If it isn’t, kill the feature. If it is, move it along to the next step.

3. Define How You Will Measure Success

A lot of things can go wrong when releasing a feature if you haven’t decided, up front, what the measurement of success of that feature is.

It’s great if the feature solves a problem that your users have, but you need to go one step further. You need to decide how you will measure success in terms of product metrics.

Before releasing the feature, you should set a hard goal. If the feature fails to meet that goal, then you know it’s a failure. At least, in terms of what you expected it to accomplish.

For example, let’s say you’ve decided to add Facebook support to your Twitter management app. The problem this is solving is that users don’t want to switch apps when managing each of their social profiles.

The predicted measurement of success could be that adding Facebook support will increase user’s session times by 50%.

Setting a benchmark like this is a good way to gauge the impact the feature has once it’s released, so it’s important to set this goal at the beginning.

Chart: time per user session in minutes

4. Mockup the Front End

The purpose of this step is to create something tangible that a user can interact with, but that can also be easily changed without much effort.

Start with a sketch, a wireframe or a pre-existing layout, just don’t create any backend functionality yet.

The goal is to create something that provides the user with a visual, tangible solution that lets them determine if it would actually solve anything for them.

Because there’s a very good chance you’ll get it wrong on the first attempt, putting only the minimal amount of effort into this step is vital.

Once you’re happy with your layout, build it in HTML/CSS, or some kind of system that is easy to adapt to user feedback.

Keep in mind this isn’t an excuse to put together something quick and dirty. This needs to look as real as possible in order to get the right kind of feedback from your users.

5. Collect User Feedback and Refine

It’s not enough to just throw this feature at a bunch of live users and ask them to play around with it. You need to set the stage properly.

The test participants need to understand two things:

  1. What your product is and why people use it

  2. What the problem is that this feature is solving

Chances are you’ll be testing on your own users, so they’ll understand the context of your product.

But if you’re using a service like UserTesting, it’s important to explain to the test participants up front exactly what your product does and the reason people are using it.

These test participants also need to understand the problem you’re solving. The best way to approach this is by telling them a simple user story leading up to the problem.

It doesn’t need to be super complicated, but it should allow them to experience the problem first hand. This way their feedback is much more genuine and valuable.

I’m not going to go into the specifics of user testing. But the overall steps here are as follows:

  1. Allow the user to experience the problem first hand (user story)

  2. Present your solution (feature) to them

  3. Collect feedback

If the overwhelming response is positive, you should move on. If it’s not, you will need to take into account everything you’ve learned, iterate your solution, and test again.

Repeat this process until you’re confident with the results, then build the functionality behind it and make it work within your product.

6. Release Partially and Collect Qualitative Feedback

At this point, you’re happy with the way the feature works. You’ve demo’d it for a bunch of your users, took their feedback into consideration, and refined it.

Great. Should you just release it and see what happens? No. Instead, release it to just a small portion of your users. This is called a partial rollout.

Choose a few customers to release the feature out to. Let some time pass, and then approach them for a formal interview and user testing session.

How is this different from the last round of testing? Because now the feature is in the real world. Your users are interacting with it (hopefully) while actually engaging with your product, not just seeing it for the first time in a user test.

While the last round of user testing was about showing them a problem and asking their feedback on your proposed solution, this round will be more a open user test.

Watch the user interact with your product freely. When they land on your new feature, ask them questions about it. Give them tasks which lead them to the original problem and see if they take advantage of the new feature to solve it.

Refine the feature with what you’ve learned from your users as you see appropriate.

7. Release Fully and Collect Quantitative Results

You’re happy. Your small portion of users you’ve released to are happy. It’s time for a full blown release.

Set the feature free and roll it out fully across your entire product. While you’re at it, send out an email blast or write a blog post up explaining the new feature.

If all goes well, this could be the game changer you’re looking for.

You’re finished right? You’ve gone through the painstaking process of refining this thing 100 times until everyone is happy. It’s finally over.

Release and move on, right?


This feature is still on trial! Worthless until proven valuable, remember?

Let the feature run for a week, or two, whatever your product needs to get a significant amount of interaction between your users and the feature.

At this point, we just want to sit back and let your users interact with it organically.

Just make sure your product is collecting metrics! Without them, there’s not even going to be a way to tell if this feature is good or bad, so be sure you have a system in place.

8. Compare Results and Decide if it Stays

The moment has arrived. Your new feature has been live for a while now. It’s time to make a decision. Does it stay, or does it go?

To start, compare the metrics from the week/month just before, and after you released. Note the impact it had on your product. What improved? Did you get an increase in users? Higher activation rates?

Most important, did it match the measurement of success you defined in step 2?

This is subjective to your product, so I can’t give you a number like “kill the product if only 1% of people are using it”. This decision needs to come from you.

If a feature isn’t succeeding, it’s hurting your UX, your budget, and your focus.

[Tweet this.]

Maybe the new feature didn’t meet your success goal, but it did bring increase your signup and activation rates. That’s still a pretty good reason to keep it around.

But if you can’t find any improvements in your metrics, then it’s vital you remove that feature. Because if it isn’t doing any good, it’s only hurting your user experience, your budget, and your team’s focus.

Feature in jail

“You can’t do this to me man! There’s guys from Friendster and MySpace in here!”

Let’s Summarize

In my experience, feature bloat is the most threatening thing to the success of your product. It affects your UX, dilutes your team’s focus, and shortens your runway by taking a serious toll on your budget.

Those are the 8 steps you need to follow to be sure the features you are adding are helping your product more than they are hurting it.

Let’s do a quick review of the steps:

  1. Start with No

  2. Identify the Problem

  3. Define Success

  4. Mockup the Front End

  5. Usertest and Refine

  6. Release Partially and Refine

  7. Release Fully

  8. Compare Metrics and Decide it’s Fate

That’s a lot to take in all at once.

To make it easier for you, I’ve put together a summarized checklist of all 8 these steps, so you can just refer back to it as you go through the process.

Click here to download a copy of the checklist.