Best Practices

How to use data to remove features from your product

Published Feb 11, 2021

With the mission to create the best possible experience for users, product managers often default to coming up with that next great feature or product functionality. While this is beneficial, one of the best things a PM can do is to retire features that aren’t being used or are no longer adding value to the product.

To do this effectively, you need deep insight into both user behavior and sentiment–which is where data comes in. When usage is extremely low or an outdated feature is constantly the subject of negative feedback, PM’s can leverage this information to make the case to leadership that a feature should be removed. Think about it: Should your team be spending engineering, design, and product management dollars on an area of your product that isn’t adding anything to your users’ experience?

Below, you’ll find five ways to take a data-driven approach to removing features from your product. 

1. Start with product usage

The first place you should start is with your product analytics. How are users navigating your application? Whether you already have a sense of a feature that is underutilized or want to see if you have any candidates for removal, it’s important to understand if users are only using the feature sporadically, or if it’s part of their usual habits and jobs to be done. 

Look for in-depth interaction over a period of time (i.e. 30-90 days)–if usage is too sporadic or not recent, your customers likely aren’t using that feature. It’s of course worth pointing out that not every feature (or product) is meant to be used every day, every week, or even every month. Be sure to keep this in mind as you examine product usage.

2. Dig into engagement

Once you have a general sense of product usage, it’s helpful to go deeper to understand how users are (or aren’t) engaging with certain areas of your product. The more data you have to back your recommendation to remove a feature, the more likely you’ll be able to do it successfully without a lot of pushback.

Look at how much time people are spending on certain pages of your application, or how often they’re clicking specific buttons. If there are any outliers on the low end, look at the paths users are taking to get to there. Do you have another feature or workflow that can solve for this in an easier way? It could be the case that these users aren’t aware of said feature or workflow. And if you’re looking at data over time and see engagement start to drop, try and find out why this might’ve happened.

3. Look at different segment of your user base

As you examine your product data, one of the most useful tactics for understanding how users are engaging with different product features and areas is segmentation. For example, if you see that a feature is popular with your small business customers but completely ignored by users from large companies, that’s a valuable signal. If your goal is to attract larger companies, you may need to prioritize different capabilities.

When looking at which segments of users access particular features, here are some ways to slice the data:

    • By company size
    • By persona (Is the intended persona accessing this feature? If not, why?)
    • By NPS response (Is the feature most commonly mentioned by NPS detractors?)
    • By role

4. Leverage feedback to understand the “why”

Equally important as the quantitative data mentioned above is qualitative data, specifically in the form of customer feedback. If you start seeing trends of lower feature usage, you need to figure out what’s going on (and why). This means talking to your colleagues in customer-facing roles (sales, customer success, etc.) and speaking with customers directly. Your customers can provide some of the most valuable data that you wouldn’t otherwise have access to–they’re the ones using your product, and also who you’re building for.

If you’re able to talk to the customers who still use a feature you’re planning to remove, you can make them aware of this change ahead of time. Try and understand how they’re using the feature and how the removal will impact their workflows, and then work with them to provide another way of accomplishing those tasks. 

5. Monitor the impact

Once you’ve said goodbye to a feature, it’s important to monitor the impact of this decision and ensure users are well-equipped to get the most value out of your product–especially any users who were using the feature that was removed. So, another way to use data is to track usage, sentiment, and feedback after you’ve removed a particular feature from the product. 

As you look at your data, here are some questions to consider:

    • What paths are users taking to accomplish the task or workflow this feature solved for?
    • Are other features seeing a spike in usage?
    • Has there been any feedback from customers about the feature removal?
    • Has our NPS increased or decreased?