May 7

Why the Concept of “Features” can be Toxic

The following insights were gleaned from a presentation at INDUSTRY given by Elizabeth Ayer, formerly of Redgate Software (now with 18F) @ElizAyer. Download Product Management TACTICS eBook for more.

We can all see a brighter future in Lean, Jobs-to-be-Done, and techniques to focus on real value and feedback loops above pumping out new capabilities. Yet many times, these efforts get dragged back to the irresistible concept of a ‘feature.’ What is it that makes features so appealing, compared to problems or outcomes?

First, let’s talk about “features.”  Whenever we think about building out our product roadmap, it certainly is easy to immediately begin thinking of new features.  Features are easy to add but have serious long-term costs in functionality. Too many features can increase sales at first, but after some time, it will force a product to collapse upon its own weight.

Rather than immediately thinking of building on your product with new features, first begin to think about desired outcomes. A good example: an output or “feature” for a government might be building more roads but the desired outcome is improving traffic flow. When we focus on a pile of features, we lose perspective on the impact our product has on the real people who use it. If you’re managing outputs, you’re judging yourself on what you produce, but not on what you’re affecting.

Avoid the Cobra Effect. In the British Raj, Delhi was overrun with snakes, so they offered a bounty for snake skins. That worked for a while, until someone discovered they could just breed cobras, kill them, and sell them to the government in order to turn a profit. The bounty system wasn’t paying off, so the government stopped it. When that happened, though, the cobra breeders simply released their snakes, making the cobra problem worse. Too many features are like a pile of snake skins. At first, they help. However, in the long term, there’s a diminishing return and even increased costs.

But this is hard, because building new features is enticing and tempting. Understand those temptations to resist them. We love how features are concrete. We love how good shipping feels. We love the control. We see features as a currency with defined price for removal. However, self-control is imperative.

But, shouldn’t we be releasing some new features at least?  Yes — but only when they’re the right features.  And you can determine this by shortening the feedback loops with the right data to get accurate perspectives about how well your product is functioning and where it really does need more iteration.

Even still, don’t overemphasise shipping, even though it feels good. Treat shipping an update like a gymnast finishing a routine. She may give a smile when she “sticks the landing,” but the real celebration comes when she and her team see the score. Celebrate the feedback instead of the shipment.

That’s working based on outcomes rather than output.

As you work this through, though, be sure to be easy on yourself with this. Outcome-based management is like exercise and eating right. It’s tough but takes constant practice and is worth it in the end. To view Elizabeth’s full presentation, visit:


You may also like

Top Product Management Resources