Building a Better Algorithm

by Margaret A. Starvish

The three largest search engines, Google, Yahoo and msn, account for the vast majority of U.S. searches, but that hasn’t stopped dozens of startups from trying to stake out their own claim to the market.

Venture capitalists also don’t appear daunted by the odds. In the first quarter of this year, they’ve poured nearly $135 million into vertical and social search startups alone, according to Dow Jones VentureOne.

Why? One reason investors are bullish on vertical search firms is that they fill needs not being met by any of the other major search engines. And there appear to be some pretty curious needs out there.

Consider some of the new niche search engines:

Reva Health Network bills itself as “a database for the medical tourism sector” and lets users comparison shop for things like heart transplants and breast reductions on several continents.

Then there’s Gloomedia, which is working to create a database,, of products from every new episode of the top 60 TV shows. Those products are available for purchase, which means users can dress themselves head to toe in the clothes Oprah wears, and watch her while sitting on the same chair she uses on TV.

For people who spend more time with their Xbox than with their remote control, there’s, a games search engine with 84,000 different sources for gaming information that got its start in Japan (where it’s the 144th most popular site, according to Alexa) and recently landed $11.9 million in funding.

Social search startups are also clamoring for investor dollars., a Colorado-based startup, secured nearly a million dollars in funding from a pool of investors, including a personal investment made by Mobius Capital’s Brad Feld. The site, posits “What if searching on the Web worked the way it did in real life?” – essentially through a complex network of personal information gathered from people you trust. Lijit translates this into a technology that “connects the dots” between personal information available on sites like Blogger, MySpace, and

Baynote bills itself as a pioneer in bringing social search to business. The company’s chief technology officer has a Ph.D. in human-computer interaction from Stanford, it got Series B funding from Disney, and customers include eBay. Baynote’s technology tracks how users get to a site like Macy’s and looks at where they go once they’re there – eventually aggregating that data so that when a consumer goes to search for a T-shirt, the search engine automatically knows if he wants one from Hanes or from Calvin Klein. According to marketing and product management director Mike Svatek, this approach has increased sales from 30 percent to 50 percent for their clients. “Our growth rate is extreme,” he adds.

Then there are the search engines that are neither vertical nor social. Mahalo, which was launched in late spring by entrepreneur Jason Calacanis, uses actual people instead of algorithms to provide results. “Maholo combines human touch and technology in a way that reduces the search term to an intelligent framework,” says former Excite CEO George Bell.

Bell’s also a fan of Powerset, the natural language search startup headed by an enthusiastic Barney Pell, who likens searching to breathing and says his technology will help people breathe easier. Powerset is still in development, though Pell hopes to launch soon, and while he’s not saying he’s looking to compete with Google directly, he did note that “people are afraid to say that their technology could scale that big. We’re not afraid to say that.”

While Pell remains one of the few people willing to make such a sweeping statement, most believe there’s money to be made in the tiny space not occupied by the big engines.

What will the market will look like a year from now? It depends on who you ask.

“I think we’ll see the emergence of some very successful vertical search engines,” says CEO and co-founder Paul Forster. “There are still a lot of categories of specialized information that people have barely begun to think about indexing and making searchable.” Forster also feels all this activity will result in a greater number of viable avenues for search marketing.

Tom Wilde, industry veteran and CEO of EveryZing (formerly PodZinger), a multimedia search engine that can find keywords in videos, is less certain: “Very few [of the new search engines] have a great insight into discovering [and] processing content, making it more usable in a really unique way.” Perhaps the most realistic view is that of Spark Capital’s Todd Dagres. When asked who he thought would succeed from the multitudes, he responded, “Whoever does a better job building a better algorithm.”



4 Responses to Building a Better Algorithm

  1. […] longscorner wrote an interesting post today on Building a Better AlgorithmHere’s a quick excerptForster also feels all this activity will result in a greater number of viable avenues for search marketing. Tom Wilde, industry veteran and CEO of EveryZing (formerly PodZinger), a multimedia search engine that can find keywords in … […]

  2. Jeff Tokarz says:

    The quest to deliver exceedingly relevant search engine results is analogous to trying to park a car in Boston. Having finally found a parking space (search results) you discover that the space is too small (search results are mediocre or irrelevant) for your vehicle.

    Search engine platforms that crawl and index searchable content are passe. Yet, when combined with natural language processing, morphological analysis, latent semantic analysis, etc. searchers get more of what they really want, and less of what they don’t.

  3. quickda says:

    […] here No Comments so far Leave a comment RSS feed for comments on this post. TrackBack URI […]

  4. I installed alexa toolbar & widegt but no result see ranking of my website here.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: