Insights: The Case for Removing Features

Editor's note: we're featuring a guest post today from Keela Robison of Urbanspoon. We regularly print guest articles from those in the Pacific Northwest high tech community; if you're a reader with something valuable to contribute, please let us know!

We've all experienced that moment when we open a familiar app or website, only to discover that a favorite feature has vanished. For me, that moment came with the discontinuation of Google Reader. † Itís easy to take feature change personally and assume that the product managers are clueless. Now, being among those who prune features, I hope to share some insight into that decision making process. Iím not the first to write about the topic of removing product features (for example, I think Lukas Mathis does an excellent job articulating a more robust version of my post). I remember reading 37signals' blog and book applauding their advice on practicing "tough love" (being willing to say "no" to customers). My conviction was further reinforced during my tenure at Amazon, by Jeff Bezosí philosophies around simplicity, and his challenge to make good decisions on behalf of customers, rather than providing more options and settings when product teams canít make up their minds.

But conventional wisdom is dangerous to apply too broadly, and this is the case for Urbanspoon. We are a unique community: we rely on customer contributions (reviews, new restaurant suggestions, etc.), and a relatively small percentage of our customers contribute the lionís share of our content. These contributors are incredibly valuable, and deeply appreciated by us. And, not coincidentally, they also tend to be our "power users" who discover and use features that the majority of customers would not.

Consequently, the decision to remove features is particularly difficult for UGC-based services such as Urbanspoon, as it risks the ire of our most valuable customers. However, I believe that the argument in favor of simplicity is compelling. As a recap for those not in the industry (and who donít spend sleepless nights worrying about "technical debt"), here are the reasons software companies should err on the side of saying no, and whenever warranted, remove features.

  • Simplified customer experience. Additional modules or menu items or settings make it difficult for new customers to find what is important. Even for experienced customers, every additional option adds "cognitive load" and makes a service feel less delightful, and more like work.
  • Reduced "load time" and increased speed of the service. Sites with fast page load have higher conversion rates and more love from Google for search ranking. Most importantly, fast sites and apps respect our customersí valuable time.

But, you may say, why not just hide the features for new/casual customers, or move them somewhere that only power users will see them? Unfortunately, doing so wouldnít accomplish these other benefits:

  • Improved stability and reliability. Features introduce bugs unless they are regularly tested. The funny thing about software is that it often is intertwined in ways that make the consequences of changes difficult to predict. A decade ago, software companies would have employed armies of QA testers to do end to end tests on an app before releasing changes. The world is different now: we leverage test automation in addition to focused QA testing, and we get to move faster and introduce more innovations to market. However, despite best efforts, unintended consequences pop up, which take time and effort to diagnose and resolve.
  • Improved ability to innovate. Features add to code (software) complexity, and thus slow our ability to build or fix what is most important
  • Ease of hiring and training new developers. Simpler code results in a lower learning curve.
  • Lowered cost of technology upgrades. When companies upgrade their underlying technology, features often need to be rebuilt. In our case, weíre transitioning our website to be responsive (so that our pages resize gracefully for tablets and phones), and this requires many if not all of our website features to be rebuilt. So, now is a great time to reassess what is needed.

Because we canít rely on the rule of thumb to follow the usage behavior of the masses, we take the time to segment our customers in order to be able to specifically focus on contributors. When a feature isnít used much by non-contributors or contributors, we remove it. At times, weíve posted notices and asked for feedback on features slated for removal from our "Primes" Ė our most active contributors. We almost always hear some complaints about the idea of removing a feature, but weíve generally found positive or indifferent response from the majority of our Primes.

Ultimately, cleaning up our interface and removing low-use features will help us be a more attractive and innovative service. We know that losing features isnít pleasant, but hope that our valued active contributors appreciate the goals we hope to achieve, and will continue to support us as we evolve our service.

Keela Robison is General Manager and Senior Vice President at Urbanspoon. Robison brings to Urbanspoon extensive experience as a seasoned consumer technology executive, leading product development and operations in diverse industries like digital media, e-commerce, mobile consumer hardware and games. Prior to joining the company, Ms. Robison held executive positions and shaped product strategy at Amazon, WhitePages, RealNetworks, T-Mobile and most recently GameHouse.


More Headlines