The challenge for Blekko is that the company must figure out exactly which category it plans to own; then work to define it. At present, the most accurate term for what Blekko does is vertical search, which has been a highly elusive category for large-scale, consumer-level entrants.
Vertical search targets specific types of content, whether by format, topic or genre. As the founder of news aggregator Topix.com, which he sold to a consortium of newspaper groups in 2005, Skrenta is no stranger to categorizing content. But Blekko is attempting something that hasn’t been done before: popularizing vertical search.
“Ideally we could be the sum of all verticals,” Skrenta said from Blekko headquarters on the California’s San Francisco Peninsula. Blekko has now been in private beta for about a month.
The irony of Blekko’s highly focused business strategy is that their straight-up search results are damn good. But in a space littered with expensive underperformers (Ask.com, Cuil, and recent casualty Yahoo), being good isn’t good enough.
“There are a lot of dead bodies on the road to search,” Skrenta said. “But there are really only two search engines left [Google and Bing]. By default, we’re the third search engine.”
With only 2000 beta invites in the wild, Blekko is taking its time in perfecting its offering before an official launch. Skrenta added that, after taking about $20 million in funding from big-name investors over the past three years, they have “plenty of gas in the tank.” They certainly have no incentive to rush things.
“If you’re not ready on Day 1, then you’re smoked,” Skrenta said, well aware that hype is a search startup’s worst enemy. At least at first, Blekko appears to be banking on high-end, hypercritical users who take search very seriously. They’re off to a good start.
The company has garnered praise for introducing “slashtags”, an evolving set of filters that effectively partition the web into categories, a.k.a. verticals, based on user input. These may be topics like /parenting or /cricket, but Blekko really gets interesting when it starts presenting the Web using fundamentally different types of searches.
For example, the engine will deliver results strictly by /rank, which Blekko openly displays. Or, users may pare down their search to results focused solely on /people. In this way, the company picks up where keyword search leaves off.
Blekko has already excited search nerds by opening up a massive amount of data explaining why results appear where they do.
As Skrenta wrote on his blog, “We don’t believe security through obscurity is the best way to drive search ranking quality forward. So we have a set of tools on blekko.com which let you understand what factors are driving our rankings, and which let you dive behind any URL or site to see what their web search footprint looks like.”
This detailed information about sites in Blekko’s crawl is a potential game-changer for publishers and webmasters, providing a wealth of SEO insight. More than that, the push towards openness could put pressure on so-called “content farms” and black hat SEO purveyors by exposing their questionable techniques to searchers.
Skrenta explained that maintaining a search engine has typically been a “closed, secretive process.”He continued: “The mentality has been: it’s so fragile that if we told you how it works it would get broken by the bad guys.”
Instead, Blekko will “let its weaknesses be moved out into the open,” he said. “Let’s get to a better, stronger place. We think that approach has a more promising endpoint.”
Within that philosophy lies the key to success or failure for the new engine, which Skrenta and his team have been building from scratch since 2007.
The company will create a community around their collaborative, vertical search engine – and in the meantime put pressure on the search giants to open up their operations.
Blekko has already come a long way just by building a search engine to scale.
“Search is really hard technically,” Skrenta said. “The first thing you have to do is copy the entire internet onto your servers.” That’s not to mention creating a continuous Web crawl, content ranking algorithms, spam filters and a natural language processing system.
As they knock the rest of the bugs out of the system, Skrenta and his team will throw themselves into the ultimate challenge: define their category and own it. Or die like all the rest.
Want a beta invite to Blekko? We have some! Send a tweet to the author @kevindgrant or write to him at firstname.lastname@example.org.