Feature Creep: Tame that Beast
Who are you, and why are you here?
Everyone knows there are three constants in life: Death, Taxes, and Feature Creep. Like a shark, a product has to keep moving forward or it dies (…right?) and the way to keep that product moving forward is, usually, to pack in more features. What drives Feature Creep, though? I would argue that there really are only two specific events that cause Feature Creep:
1: The product is a hit. It has focus, clarity; it’s so easy to use. In fact it works so well, that the next version of the product should do this, and this and this…
2: The product is not a hit. In an attempt to build value, add features.
That Enigmatic Mandate
Admittedly, that sounds awfully deterministic. But, really, designers have long had to wrestle with the mandate of incorporating more features into a product and making it easier to use. It is not easy, there is no such thing as “no design” and good design doesn’t just happen. What does that mean for designers of map-based data visualizations? Plenty. I’ll attempt to lay out strategy for pulling it off a bit farther down, in the meantime here are some of the arguments pro-ing and con-ing Feature Creep.
Less is More
Diminishing Marginal Utility
Some would argue that every new feature diminishes the marginal utility of the feature set, watering down the perceived value of every other feature.
Negative Marginal Utility
Further, some would argue that a tipping point can be reached where utility is not just diminished, but reversed. I have run across some online map applications that present me with a big set of iconized tool buttons, none with any kind of text description, and placed in no discernable order across a catch-all toolbar at the top. I can tell you that I am not just less likely to click on any of those buttons, but I would just assume leave that application altogether. There have been some interesting studies that quantify the walk-away effect in the worlds of Jam, and 401Ks.
In Malcom Gladwell’s book, Blink, he describes how Sheena Iyengar set up an experiment to test the buying likeliness of shoppers who were offered 24 different types of jam versus buyers who had 6 types of jam from which to choose. The group with 6 jam options were ten-times more likely to make a purchase. Similarly, 401K participation rates drop an average of 2% for every ten plan options offered. When asked to choose from more and more options, people will increasingly choose nothing. The unfortunate commuters who are subject to this stoplight, however, probably don’t have that luxury.
A revealing series of studies in the Harvard Business Review by Roland T. Rust, Debora Viana Thompson, and Rebecca W. Hamilton describe this phenomenon. “Manufacturers that increase a product’s capability–the number of useful functions it can perform–at the expense of its usability are exposing their customers to feature fatigue.” They found that “even though consumers know that products with more features are harder to use, they initially choose high-feature models. They also pile on more features when given the chance to customize a product for their needs. Once consumers have actually worked with a product, however, usability starts to matter more to them than capability.” That knife below is an actual Wenger Swiss Army Knife. It costs $1,200 and was out of stock at the time that I wrote this.
No, More is More
In his essay, Simple Minds, Paul Kedrosky argues that the key is to “have more features and more information in ways that are less intrusive and more carefully prioritized.” Yes! The push for simplicity should not confuse simple with simple-to-use. Simple implies a relatively lame feature set, while simple-to-use is the result of some kind of black magic that makes a powerful product practical. Eistein probably would have agreed… "Make everything as simple as possible, but not simpler."
The Upside of Feature Creep
Well…you get paid. So the shark isn’t moving forward just to stay alive, it’s moving forward to make a living! Granted, poorly integrating the new features will kill a product in the long run, but well-integrated relevant features ensure its future -and yours. Additionally, carefully selected feature requests from clients push the product to the bleeding edge. That’s direct, focused, market-driven research! What better way to find out what features will enhance your product than to listen to your clients?
The flip-side of this free market research is that you’ll have to sort out what feature requests are awesome, and which ones are pet features or misguided ideas that will have no larger audience. I have found that there are three common methods for prioritizing feature requests. The three Ps:
Price it. This can be the best method for weeding out pet features. When presented an honest estimate of what a pet feature will cost, often it is rightly pruned. Remember how much the Giant Swiss Army Knife cost (though it was sold out…)? "The cheapest, fastest, and most reliable components are those that aren’t there." — Gordon Bell
Phase it. Discuss with the client which features are more urgent than others. Ease users into the product’s feature set through iterative releases; they are more likely to respond well to the product if they are immersed into the capabilities step-by-step. Sometimes a feature may benefit from the user’s preexisting familiarity with the product. These dependent features are great candidates for future releases of the product.
Push it. Don’t be afraid to say no. Sometimes a feature has no audience and no amount of crafty design can prevent it from weakening the product. Policing features is a must in order to keep the product relevant. You may just have to push back, like the subjects in the feature fatigue study wish they had. You are the gatekeeper and your clients will thank you in the future with their continued business.
Have Your Cake and Eat it Too
Only expose a feature to the front end (provide some widget to the user) if it absolutely needs to be there. Some settings are best left assumed. Remember the Seinfeld episode where Jerry’s assistant makes his trip a nightmare by asking him inane amounts of questions about his preferences in her effort to ensure his satisfaction? Riiight. It’s funny because it’s true.
Group like tools. Categorization is the most important method to handle an interface with more than a few items. This is the most important task after deciding what goes into the interface in the first place. Carefully consider what each feature is for and assign it a feature family. Show and Hide. Not everything needs to be visible all the time. Be very lean with your default settings and nest your features in a manner that they are easily accessed. Web users navigate by scent, so name your category, and nest those features. In an article in Fast Company, Marissa Mayer, Google’s home page gatekeeper, likens Google’s functionality to a really complicated Swiss Army Knife. "…the home page is our way of approaching it closed. It’s simple, it’s elegant, you can slip it in your pocket, but it’s got the great doodad when you need it. A lot of our competitors are like a Swiss Army knife open–and that can be intimidating and occasionally harmful." I doubt she had the Wenger Giant in mind when she offered that quote, but only Marissa knows for sure.
Take a look at Microsoft Word. The product has steadily grown its offering of features, but Word 2007 focused on a usable interface as its latest feature enhancement. Check out line one of their product description: "Create and share great-looking documents by combining a comprehensive set of writing tools with an easy-to-use interface." The latest feature is an enhances method of prioritizing and organizing the universe of features available in the ubiquitous word processor.
Word Then (from Jeff Atwood’s blog)
Word Now (from the Word website)
Design is a direct result of understanding a client’s pain point then being able to draw a picture of the solution. "Good design is clear thinking made visible." Edward Tufte said that -then I started saying it a lot, too. Because he’s right. At IDV, I’ll almost never pick up so much as a pencil before I have a clear idea of what it is that the client needs. This isn’t to say that good solutions are dictated. They evolve as the result of clear communication and careful consideration. It is key to recognize that continued requests for new features are not a nuisance, but a valuable means of honing an elegant solution. Feature Creep is not just unavoidable; it is necessary.