A former Google search scientist and an ex-Cisco product exec could have had their pick of startups and they chose to replace one of the Web’s dark arts: search engine optimization.BloomReach came out of stealth mode yesterday with a healthy list of 70 big-name customers and a serious pedigree of talent from Cisco, Google, Facebook and others.
BloomReach has collected $16 million so far in two rounds of funding, lead by Bain Capital Ventures and Lightspeed Venture Partners, respectively,
Co-founder and CEO Raj De Datta would quibble with SEO label. SEO evokes a certain smoke-and-mirrors art where companies use things like meta tags, title tags and inbound links to make their web sites surface around specific keywords.
BloomReach doesn’t do any of that, De Datta told Business Insider.
It uses big data and machine-learning algorithms to help online retailers put more of their products in front of shoppers via Google, Facebook, or social media sites.
Even the biggest, best, most optimised websites see most of their traffic only going to about 25 per cent of their web pages, De Datta says. Using big data analytics, BloomReach figures out a searcher’s intent and can help a website deliver a more relevant page.
It wasn’t an easy problem to tackle. It took years.
In late 2008, De Datta was urged to talk to Internet marketeers by the VC he was working with at the time, Justin Caldbeck at Bain.
He discovered that enterprises were “spending enormous amounts of money on online marketing,” he said, and yet they were still struggling with how to get people to find their sites. The SEO consulting industry had arisen but, “this smelled to me like the kind of problem where a more structured, more data-oriented approach would have a bigger impact.”
He didn’t know how to go about that until he met Ashutosh Garg.
Garg had been a staff scientist at Google for over four years where he worked on some of Google’s most important search services. As a PhD and former IBM researcher, Garg is well known among the machine-learning elite.
Garg had left Google a few months earlier thinking he was going to build a better search engine. The response he was getting was “are you crazy?”
De Datta and Garg realised they wanted to solve two sides of the same problem — making it easier for people to find stuff on the Web.
In February, 2009, Bain “Made a big bet,” De Datta recalls. “It was really just the two of us and a PowerPoint presentation …but they gave us $5 million”
Instead of using that $5 million to staff up, the two cautiously hired a handful of machine-learning PhDs and left enough money in the bank so they could “afford to fail a couple of times before we got it right,” De Datta says.
Three years later they are on their way. They now have three services covering search, advertising, and social media. They employ about 60 people and their customer list includes the likes of Orbitz and Williams Sonoma.
They are so confident that they avoided a straight subscription model and instead adopted a pay for performance model, too. Their customers pay only for the incremental traffic the service delivers.