The Goldilocks Algorithms: Someone's Been Peeking in My Head
When MAYA’s Goldilocks Algorithms present information that’s “just right” at every level of interaction, users feel like the system is not only responsive, but also anticipates their desires. That’s truly intuitive design.
Users of complex interactive systems often feel like they have little control over the amount of information coming at them.
It either overwhelms them (like drinking from a fire hose), or eludes them (like looking for water in a desert). What they crave are interfaces that give them what they want the moment they want it. (It’s like a thoughtful bartender who pours your favorite martini just when you’re thinking, “I’d like a drink . . .”)
MAYA has become very good at designing mechanisms that mediate between huge stores or flows of information and the desires of users as they make their way through that information. We’ve dubbed one of those mechanisms, “The Goldilocks Algorithms.”
Like Dave Bishop’s Goldilocks Principle, the Goldilocks Algorithms help MAYA tame complexity. But instead of setting the goal, these algorithms help us achieve it by determining exactly how much information to display at specific points of use.
We’ve used the algorithms for years, even in projects that predate widespread use of the Internet (see Ameritech | The Neighborhood). But the effects of the algorithms are especially obvious with the interactive Greenmap that we helped create for 3 Rivers Connect.
Because the interface of this map gives users great freedom in how they pan, zoom, and make selections, we had to apply the Goldilocks Algorithms to constantly assess which features are appropriate to show on the map at any scale.
Not too much: clutter control
Depending on the scale and the number of items a user has chosen to view on the map, the display could become cluttered with many overlapping items. Think of displaying all of the restaurants and banks in a major city. We control the clutter by aggregating the information in the densely overlapped areas and allowing the user to see the details behind the aggregation(s) by zooming in. (See Figure 1 for an example of how clutter can hurt all kinds of interfaces, not just maps.)
Not too little: sparseness control
On the other hand, say the user hasn’t selected any features, or there isn’t much to show in the categories the user has chosen to view on the map — a common condition at very large scales. Instead of displaying a map devoid of features, we fill the voids with the most salient, unsolicited information. This display of unsolicited information encourages the desired behaviors of exploration and discovery by exposing content that the user might never see otherwise. (The sparseness of the interface in Figure 2 forces users to work hard to figure out what to do.)
Just right
There is some subtlety in how MAYA fine-tunes these algorithms to get the user experience right, such as decluttering with the fewest number of aggregates, yet indicating appropriate density, and figuring out what unsolicited information is the most salient. (Although it doesn’t use our Goldilocks Algorithms, appropriate density helps to make the Web-based content-management system in Figure 3 easy to use, even for people who aren’t technology experts.)
MAYA Design is developing a geobrowser: a way of intuitively searching, browsing, and displaying geo-coded information through a map-like interface. Even though the interface must accommodate more complex and diverse data sets that are many orders of magnitude larger than the Greenmap, the Goldilocks Algorithms make it just as easy to use.
A usability goal for any product should be an interactive experience that is as welcome as that thoughtful bartender — intuitive, natural, and just right.