It was three years ago today that a handful of guys struggling with a stinker of a product had an epiphany and three weeks later transformed it into what could be what saved Twitter.
April 11, 2008 Jay Virdy and team were working in a small office in northern Virginia on a search engine that was going exactly nowhere.
“We didn’t get any traction with our Review Search Engine angle using cutting-edge sentiment analysis technology. We were solving too complex of a problem instead of solving something simple,” said Virdy, a former search guru at AOL. “Our chief scientist, Abdur Chowdhury, joked that we built a Summize Flux Capacitor.” Ahh, geek time machine humour… Actually, that would have been more impressive than their existing tool.
So how did they go from fictional movie conceits to launching a so-called “conversation engine”? Let’s just say desperation can do wonders — generally you either break through to new heights or ya flame out. Summize broke through.
Said Virdy: “After being backed into a corner I remembered what Brad Burnham of Union Square Venture had advised me to do: ‘Build something simple. Let the market pull you in.’ “
“Also, I was aware that Twitter was working on search.” TechCrunch wrote about it at the time: http://techcrunch.com/2007/09/24/real-time-search-now-at-twitter/
However, “The tipping point for me was this tweet from @Loic [Loic Le Meur, founder of Seesmic] that convinced me to pivot from a review search engine to a real-time conversational search engine:”
March 20, 2008: after Terraminds now it is Tweetscan that takes forever to result my query. We really need good Twitter search. (Actual Tweet is here.)
OK, so now there was a problem identified and inspiration to solve it. What next? Virdy humorously notes that the goal at launch was “about survival.”
“We had run out of money, we had run out of other options. It was our Hail Mary pass moment.”
But it wasn’t without some additional guidance. Virdy looked to Stephen Blank’s book Four Steps to the Epiphany to identify a unique opportunity that existed now that Twitter gaining momentum with the early adopter crowd.
“The entire Summize team pulled this together starting with Matt Sanford who I think was our MVP. I knew Matt could figure out a way to get the real-time Tweets or, if necessary, aggressively crawl Twitter.com as a last resort, Virdy said. “I knew Mike Hayes could easily write the code to extract real-time trends. Eric Jensen and Abdur Chowdhury could harness MySQL to implement a real-time search engine, and Greg Pass could bring all the data together in a beautiful user interface.”
To fine tune the “it” thing his new product would have Virdy said he went back to Burnham’s comment in one of their meetings with USV. “He said a product needed a small twist to gain traction,” Virdy said. “I believe Trending Topics was our small twist that separated us from other Twitter search products.”
Trending Topics at the time was especially helpful for new users trying to understand what the heck Twitter was.
“We also had other innovations like shortened URL previews, language translation, conversation threading, that the early adopters appreciated,” said Virdy.
Three weeks later the new Summize was launched. Three months after that Twitter snapped them up. Three years after that Twitter search is a core utility online, and trending topics? Well, I won’t insult your intelligence.
So what’s Virdy’s next act? He’s leasing his skills to the government as a Gov 2.0 consultant and is an angel investor and advisor to small startups in Northern Virginia. Follow him @jayvirdy.