When I asked Ask.com president Doug Leeds about recent reports on the search engine’s layoffs, he had an immediate answer. “The headlines are out of whack,” he said.
The company has been hammered by critics in recent days. After Ask cut 130 engineer jobs Tuesday, Bloomberg dramatically proclaimed: “Ask.com Surrenders to Google.”
For Leeds, the decision to slash Ask’s workforce was unfortunate, but ultimately healthy. It was a necessary sacrifice to refocus the company on what it does best: question & answer search, with a human-assisted approach.
How did Ask make the decision to move away from algorithmic search, the engine competitors Bing, Google, and Yahoo rely on?
“Really, we’ve been around longer in fact than Google, and started out as a Q&A service,” explains Leeds. True, then-named Ask Jeeves was founded several years before Google, but it was Larry Page’s PageRank algorithm, developed at Stanford in 1998, that truly changed the industry. (Ask integrated similar technology after it acquired Teoma in 2001.)
“Our innovations in algorithmic search were either being matched by other companies or copied by those companies, particularly Google: the more their share grew and grew and grew, the more our innovations became their innovations,” Leeds continues. “You have this fawning press corps that gives Google credit for having invented anything that they do, even if it was invented somewhere else.”
In Leeds’ telling, it reached a point where most keyword-based search engines were producing very similar results. He realized the war of market share in algorithmic search wasn’t worth fighting.
“Everyone has the exact same, vanilla results–and that battle has been won,” he argues. “If you’re going to compete in search, you’re going to have to create a completely leap-frogging technology, which, for some reason, Google can’t copy. Well, there isn’t a leap-frogging technology that Google can’t copy. So you have to do something that Google wouldn’t want to copy, or something that Google isn’t interested in copying.”
Leeds believes Ask has found an answer for Google in human-assisted search. The engine’s focus is now two-fold: crawling the Web for questions that have been answered, and creating answers for questions that haven’t. The company’s 250 employees are laser-focused on this goal of what Leeds calls “one click” answers, and on creating a system that ranks the results intelligently. (Head here for a detailed run-down of the engine’s features.) So if a user asks “Why is the Golden Gate Bridge red?,” the engine provides the following answer at the top of the page:
While most people think of this iconic bridge as red, its true color is “International Orange.” Architect Irving Morrow chose the color for its visibility and attractive blend with the surroundings. Source: goldengatebridge.org.
It’s an arduous process, but Leeds says Ask is already seeing results. Since focusing the engine on human-assisted search, Ask has gone from 25% to 30% of queries being entered in the form of a question; this summer, when the company redesigned the page around its Q&A service, that figure shot up to 46%. Users who experience the Q&A service, Leeds explains, “came back more often in a month, and more months in a row.”
Currently, around 60% of the questions asked have answers; Leeds’ aim is to reach 90%, and he is expanding the human-assisted search team to do so.
“No parents are putting a report card on the refrigerator door if their kids are only getting a 60% grade,” Leeds says. “But we’re already seeing a real change in behavior. Before, we were looking at data showing that if you make your algorithmic results better, you will get better user loyalty. But it was much harder to move the needle. It took months and months, with dozens if not hundreds of engineers, and tens of thousands of servers, just to move that needle a little bit. We realized we could move the needle just as much, really more, by putting a lot fewer resources into a very different platform.”
For Ask, which has seen its algorithmic search market share crumble over the years, there was no reason to stay the course.
“Yahoo faced a similar issue as us: People just weren’t coming to Yahoo to use keyword search,” Leeds explains. “Microsoft can spend hundreds of millions of dollars marketing Bing–we’ve been there too. Look, I like to say that ‘Ask’ was a verb before ‘Google,’ but in our culture now, Google is a verb that means ‘to search.'”
Leeds expects dramatic differentiation in search to be an industry trend. That’s why Google has rolled out Instant; why Yahoo and AOL have shifted toward becoming news portals; and why Microsoft’s Bing has teamed up with Facebook for social search. Users are impatient for a better system than algorithmic search can offer. As Leeds jokes: “What’s on page three or four? Who cares? Nobody is even going to get that far!”
“While Google may now be synonymous with keyword search, we are synonymous with questions and answers,” the Ask.com head says. “You could say we had a shoe store. But when you walked inside, you couldn’t find shoes anywhere. All we’re doing now is putting the shoes up front, so people can find them.”
“To the extent we can do that, we’re going to start selling shoes again.”