Twitter’s foray into search (through last year’s acquisition of Summize) has been commented on by every pundit under the sun. Twitter Search is has proved the benefits of real-time search — namely quicker access to feedback, which gives the ability to respond and steer the person’s experience.
Businesses have found this useful, the limitations of real-time search (RTS) have kept it primarily out of consumer search habits. Twitter today could make a lot of money by sitting as the platform for feedback and service between businesses and consumers, but to justify their recently rumored $1B valuation, you can bet solving the limitations of real-time search (RTS) as it applies to consumer search is on their radar.
The big problems with RTS revolve around the inability to distinguish noise from signal in the short term. Just being the most recent doesn’t make you the most interesting, but it does make you more interesting. The hard problem to solve here is how to factor timeliness into the algorithm for search relevancy. Twitter Search currently understands timeliness, and understands the basics of how people are voting with their actions (although, trending topics is just scraping the top layer of something that needs rich click AND publish data to be interesting), but it doesn’t quite have the larger search algorithm game figured out. Twitter has to internally solve the problem of bringing in more standard search knowledge and expertise.
Bit.ly, a startup that shortens links for use in micro-updates, has developed an incredible (and timely) database on how many people are posting links, how many of those people are unique or simply reposting from someone else, and how many people are clicking them. That data could be very interesting to Twitter as a way to dig deeper on trending, but it could be even more interesting to someone who is trying to work real-time information into an existing search algorithm. Do we know any companies that are very concerned about making sure they are on top of the next innovation in search? Google comes to mind; Microsoft Bing should be paying attention too.
Twitter could benefit from bringing bit.ly in house, but I don’t think they’d benefit as immediately from acquiring bit.ly as Google or Bing would. Not that bit.ly has to sell, but, they’ve only raised 2M, could likely get a large sticker price, and could get to help reshape a search service relied on by hundreds of millions of people. If I were BD at GOOG or MSFT, I’d be starting conversations.
Bonus: Twitter, I’m not just dishing out free BD advice to the big guys, you get some too! Buy or build a service similar to Blippr and automatically give feedback on products based on what people are saying on your service, then, license this rich “micro-review” data to companies like Amazon and RichRelevance.
Come see me speak, Oct. 6th in Sunnyvale