By nature, people require some semblance of control in their lives; without it we are prone to depression, but with it we are happier and live longer. This is supported by Langer and Rodin’s nursing home study in which all residents in the study received the same healthcare benefits, but those who were given choice and personal responsibility over their benefits experienced greater joy and had lower mortality rates.
Our choices give us identity and emphasize our diversity. Through personal investment, they drive loyalty and improve emotional ties. These benefits of choice can carry over to our digital experiences, but careful consideration should be exercised when allowing users choice and customization because of the many associated risks.
Balance
One extremely important consideration is how much control to give a user. There are heavy implications of going too far in either direction.
If the level of control is too little, users’ sense of ownership of the application will be lessened. When confronted with a difficult task within the application, if they have no ability to control or avoid the situation they are likely to get frustrated and lose motivation. This sort of behavior is shown in Seligman and Maier’s learned helplessness experiment where subjects who could not affect or avoid adverse situations showed signs of resignation and clinical depression.
On the other hand, if users are given too many things to consider, they experience an unacceptable cognitive load and become stressed. This leads to paralysis and poor decisions. Sheena Lyengar’s jam study showed subjects becoming paralyzed and refusing to buy any jam when presented with more than six options. Shiv and Fedorikihin conducted a study in which people made emotional (as opposed to rational) decisions when overloaded with too much mental processing. In that study, subjects whose mental resources were heavily taxed in an attempt to remember a seven-digit number would opt for an unhealthy snack that offered immediate comfort and didn’t require additional thought.
Unfortunately the challenge of figuring out the proper amount of control is compounded by the fact that people are diverse and require different amounts of control depending upon their personalities, workflows, current tasks, and typical uses of an application.
Because of this range, it is wise to consider a tiered approach to customization. Users should be presented with a set of “smart” defaults that will be optimal for the majority of people. Then they should be given a quick and simple way to make broad sweeping changes, and from there they should be able to perform additional fine-tuning at an advanced level. This is an approach similar to Windows’ themes and Adobe’s panel sets. Novice users and others in a hurry can get to work immediately. Casual users can do simple customizations and explore further adjustments. Advanced users can alter the interface quickly for their current tasks, and hone in on setups that work best for them.
Active Versus Passive Adjustments
Sometimes adjustments to an application occur without the user consciously triggering them. Depending upon the scope and nature of such adjustments, they can be good or very confusing. Even if they are helpful, passive adjustments may fail to elicit the kind of emotional bond that active ones do since there is less personal investment involved.
These adjustments are confusing and frustrating when they are triggered too easily or change too much. This often occurs when hot keys radically alter an interface. While hot keys are extremely useful to expert users, they can be rather jarring for novice users. The “F” and “Tab” keys in Adobe Photoshop are great examples of this. In a single stroke users can either go to full screen mode or turn off all of their panels. Photoshop attracts a wide range of user types and many don’t know about these keys. Autodesk’s 3ds Max has a similar command to launch “expert mode.” The difference, however, is that their command requires two keys and has a clear indicator that you have entered expert mode that also offers an easy way back.
Passive adjustments are nice when the application picks up on the trends in the user’s behavior over time and adjusts itself in accordance. An example is when recent tasks and often-used commands are surfaced in the interface, like the in the Windows Start menu. This kind of customization is often efficient because the system is detecting the trends as opposed to being based on subjective choices. These changes should be predictable, reversible, and somewhat limited in nature because extensive automated changes can be jarring to the user and cause confusion.
Range
When deciding how much customization to offer users, consider the purpose and context of an experience. Some applications are quite complex and require users to have a large amount of control while others are extremely basic in scope and customization would only be feature bloat.
For example, a simple notepad application is most often used for taking quick notes. One of its primary benefits is its ability to load fast and allow the user to type notes quickly. Adding a lot of features to customize this could defeat the application’s purpose. Labeling and grouping notes would be advantageous, but should be easy to do afterward.
In cases where a lot of options are necessary, it is best to consider presenting them in consumable chunks by grouping and progressively disclosing them. Old VCRs did this by exposing the most common basic controls to all users while grouping and hiding advanced ones under a flap. In a similar fashion, many programs have their options displayed in tabs with a “general” one coming first and an “advanced” one at the end. This approach helps prevent novice and casual users from feeling overwhelmed while also facilitating advanced users in finding what they need.
Facilitation
When users are offered the ability to control experiences, they are also given the ability to fail at doing so. As mentioned earlier it’s a good idea to make changes predictable and easy to undo, but it’s even better to prevent that failure in the first place. Therefore designers need to consider how to inform the user as well as prevent them from “failure.”
Tool-tips and other forms of contextual help can assist the user in understanding the changes they are about to implement or have the ability to control. These are often seen when an application is opened for the first time, or is upgraded to a new version with features that need explanation. This method was used to help explain tabs when they first started showing in browsers.
Online customizers are notorious for allowing their users to fail. For example, it is far too easy to create a pair of “clown shoes” with some of the shoe customization products on the Internet, and an excess of customization was a contributor to the downfall of MySpace. These customizers would better serve their users if they assisted them in pairing proper colors through suggested palettes. Sites such as Kuler offer an elegant solution by receiving a color input and returning various swatch pairings that the user can choose from. Then if the user wishes to deviate from there, it allows them to do so.
Darren Solomon’s In B Flat demonstrates the idea of fail-proof control. On his site a user is presented with 30 YouTube videos with content ranging from instrumentals to spoken word. The user then can play any number of videos at any time in any order and will always end up with an appealing acoustical experience, one that is so pleasing and unique that it invokes a sense of pride of authorship.
Conclusion
Control is important to us all and it is necessary to consider when designing a digital experience. While it may be tempting to assume it is easier to give a user control over an interface, it often means more work in order to assure the appropriate amount, as well as sufficiently facilitate and manage that customization process.
Finding the right balance can be difficult but is definitely worth it if done well. Customization allows users to enjoy the experience more while probably also overlooking any of its minor flaws. If users can and do customize an interface, they will interact with it more and have greater investment in it, creating the positive emotions of pride and ownership.
