type
words
for
people
Redesigning Search
August 2009
Search engines look and feel pretty much the same today as they did ten years ago, when Google first rose to prominence. Why is that? Is it because we got everything right on the first try? Or is it because we just lack imagination?
A while ago I attended a Q&A led by Udi Manber, the VP Engineering in charge of Google’s core search technology. Someone (not me) asked him why Google’s search interface had undergone so little change to date, and whether any significant changes were planned for the future. Manber seemed to think these were silly questions. First of all, he said, the search interface has changed a great deal even in the last two or three years; and second, it’s impossible to fundamentally improve on lists of blue links over a white background. Apparently, Google’s interface is close to perfect, and all that’s left is some subtle tweaking.
I guess this is the kind of complacency that comes with creating the most successful thing in the history of the internet: even your flaws start to look like strengths. Seemingly well aware that they had no sense of design, Google minimized the impact of their greatest weakness by keeping things as sparse as possible1 — and their success convinced them and everyone else that sparseness and a designless sort of design are intrinsic to any good search engine.
Google’s engineer-dominated, data-fanatical culture frustrated its senior visual designer, Doug Bowman, to the point of leaving the company in March to become Twitter’s creative director. In a piece titled Goodbye, Google, he said this:
When I joined Google as its first visual designer, the company was already seven years old. Seven years is a long time to run a company without a classically trained designer. Google had plenty of designers on staff then, but most of them had backgrounds in CS or HCI. And none of them were in high-up, respected leadership positions. Without a person at (or near) the helm who thoroughly understands the principles and elements of Design, a company eventually runs out of reasons for design decisions. With every new design decision, critics cry foul. Without conviction, doubt creeps in. Instincts fail. ‘Is this the right move?’ When a company is filled with engineers, it turns to engineering to solve problems. Reduce each decision to a simple logic problem. Remove all subjectivity and just look at the data. Data in your favor? Ok, launch it. Data shows negative effects? Back to the drawing board. And that data eventually becomes a crutch for every decision, paralyzing the company and preventing it from making any daring design decisions.
Interfaces matter. The way we perceive information is shaped by the medium through which it moves; similarly, the precise way in which we interact with a piece of technology influences what we want that technology to do for us. A Google search results page is in no way a pleasant place to be — nobody wants to spend much time there. You only go there to get sent away to someplace else. When you do a search, you want it to be over as quickly as possible. Google’s interface has trained us to think like this. To a significant degree, the interface has ended up defining what search is and isn’t.
Looking up movie showtimes, finding the number for a pizza place, making sure Jeff Goldblum is still alive — these are the kinds of searches that are well suited to an uninviting interface, where the only pleasurable part of the experience comes at the end, when you locate the specific item you’re looking for. These are “directed” searches: you start with some idea of the piece of information you need to find, and when you’ve found it, the search is over.
But it’s possible to conceive of other ways of searching, and some of these are already beginning to emerge outside of Google. Twitter Search is still search, but it’s not directed: instead of starting with a prefigured goal, you start with a topic, and you search to see what the (Twitter-using) world is saying about that topic. The search ends when you get bored.
It’s helpful to imagine searches as existing on a spectrum between these two extremes. At one end, we hunt: there’s a concrete fact we want, and we search until it’s captured. At the other end, we gather: we’re looking to collect information about a topic, for understanding or entertainment. Most searches today probably fall somewhere in between: we search because we need help making a decision or answering a question; we have a sense of where we’re going but we might be willing to explore a little.
So, if we were to design a search interface that accommodated a broader spectrum of search modes, what would it look like? It would have to offer a highly flexible user experience that’s enjoyable enough to lend itself to undirected exploration without being inefficient or overwhelmingly playful; it would have to be able to pull information from a variety of sources, through a variety of filters; it would have to be either extremely intelligent or effortlessly customizable, so that it knows (or we can easily tell it) what kind of information we want and where we want to get it from. It might look like this:
At WebMynd we’ve been developing a new search interface framework called Delve. In addition to presenting Google search results, Delve interfaces draw information from a customizable set of sources, which can include sites like Twitter, YouTube and Wikipedia. A browser add-on applies your preferred interface to your Google searches, so you’re not moving to a new search engine — you’re just augmenting the one you already use.
Shown above is one version of our first experimental interface, Phoenix, which we've just released. It’s a rough draft of a solution to a seldom-addressed problem: the lack of real innovation in search interaction design. Phoenix is pretty, but that’s not really the point. What matters is that it suggests new kinds of functionality with a new kind of search experience.
Search is still the most important part of the online experience. The interfaces to our search engines are effectively our interfaces to the web itself, and they have a profound influence on how we interact with the information around us.
Google’s inability to grasp the importance of design has loaded us with some serious mental inertia: after ten years of the same interface, it’s a struggle to switch to something we’re not used to. But it might be worth it. We’re still in the early days of the web, and I have a hard time believing blue links on a white background are as good as it gets.
Imran Zaidi (that's me — hi) is the design lead at WebMynd, a Y Combinator startup based in San Francisco. This is his (my) weblog.