Improve User Experience by Reducing Features

A common mistake when designing or changing a user experience (UX) is to overload it with features in an attempt to address all possible needs. This is a phenomenon termed ‘featuritis’ by Gartner.

Magnus Revang, research director at Gartner, is speaking at the Gartner Application Architecture, Development & Integration Summit 2016 in London today, and he said that often the UX for applications or websites is designed or purchased by people who aren’t going to use it themselves. This can make it difficult to evaluate in terms of quality.

“ Find the right product before making the product right”

"Creating a user experience can become a case of ticking off a list of business requirements, adding one feature after another without considering what features are most important to users," Mr. Revang said.

Gartner research director Magnus Revang explains 'Featuritis'  

The psychology of choices

Why is having lots of features a bad thing? It comes down to well-documented human psychology, referred to as ‘the paradox of choice’ by psychologist Barry Schwartz. Beyond a certain point, too much choice paralyses our decision making and makes us more dissatisfied. A well-known example of this phenomenon was an experiment in a supermarket, where researchers found that consumers were more likely to buy jam when offered a choice of six varieties instead of 24.

The usual cause of featuritis is that the people in charge of the project don’t spend enough time doing research and fostering insight to create empathy with users. They need to find the right product before making the product right. Ideally around 20 percent of time will be spent evenly across the project, researching user needs and gradually transitioning to validation of the solution.

Curing featuritis

It’s important to understand that all software reaches a point when the next added feature degrades the UX for all current and future users. A lot of software has gone beyond this point already, and both users and developers would benefit from some streamlining. Mr. Revang outlined thee three strategies to reduce features from an existing UX, rather than designing one from scratch.

• Removal — This is the most obvious path, but not an easy one. Removing features is not easy because there will always be someone (or something) that depends on a certain feature. The removal process involves identifying all the dependencies — both human and systemic — and slowly phasing features out, replacing their functionality in better solutions as necessary.
• Compartmentalization — This method involves creating several separate solutions, with fewer features in each, to replace one feature-heavy UX. For example you may have a complex user interface that works for a small group of super users, who need all its functionality, but that is far too complex for the vast majority of users who use only one or two features. Duplicating the most commonly used features in a separate app will reduce both complexity and training needs, and will allow different development teams to focus on the varying needs of different user groups.
• Reframing — This consists of removing a feature from the front end of an application and incorporating it into the process of using the application, so that it appears only when relevant. Advanced search function is a good example. Instead of having a separate feature and control for advanced search, it makes sense to offer the advanced search fields only after a search has taken place to filter the existing results. This presents the user with a choice when it is relevant.

Reducing features in this way should ultimately improve the UX while reducing the development burden for designers, allowing them to focus on quality within the existing features.