Destruction Keeps You Young: Why Removing Product Features Can Be as Important as Adding Them

Kevin Davis

The best part about working in software? We get to invent the future out of nothing. It’s incredibly rewarding to see something you dreamed up a few days ago being used and appreciated by people all around the world.

However, it doesn’t always work out that way. Sometimes people find your idea horrifically annoying, sometimes no one can understand what you do, and sometimes it just doesn’t catch on like you’d hoped it would.

At Yammer, we run a lot of experiments, and a lot of them don’t ship because they haven’t proven to have a positive impact on how people engage with the product. But despite really good discipline, sometimes a stinker of a feature makes its way into the product.

A while back, you could send and receive messages in Yammer through SMS. When my colleague looked into whether people were using this, it had steadily declined to the point that it was being used by less than 10,000 active users — a really small percentage of Yammer’s overall usage. We expected that most people using it had either switched to our since-released mobile apps, or had left the service after we’d provided them with a pretty crummy experience via SMS: You had to deal with a vendor, there was a litany of support issues with all the different providers, and SMS couldn’t keep pace with all the different things that were happening on Yammer.

Ceteris paribus, simpler products win.

Your customers’ top concerns have nothing to do with your software. If you’re writing software for businesses, your customers care about making sure they can make payroll, whether their new hire is going to work out, or what will happen if no one wants to buy their new product. They could give a shit that you added a new admin toggle. If your customers perceive your product to be confusing or “not for them,” you’re sunk — they have decisions to make all day, and your software shouldn’t be on their list of things to figure out. People are tired of making decisions, and if using your stuff is something your customers dread, they’ll beat a path to what ever shows up on their first web search.

Maintaining features nobody uses is incredibly expensive from a customer standpoint: every moment you spend on a feature people don’t use is a moment you’re not spending on a feature they do use. The other cost is something Kris Gale goes into at length—keeping stuff in the product that isn’t delivering customer value is a huge engineering expense. Every change that touches that area of code, every new hire that has to ramp up, every sync and every deploy, you pay a hefty price for features that aren’t being used.

The less time you can avoid being distracted by your mistakes, the more you can focus on the things that matter.

In a companion post, Yammer’s Senior Manager of Enterprise Strategy Ryan Braastad talks about how we remove features here at Yammer, where changes impact thousands of companies and millions of users.

Kevin Davis is a Product Manager at Yammer. He loves white water rafting, pizza, food, sharks, whiskey, ultimate frisbee, and being Canadian.