Micro-blogging phenomenon Twitter Inc. hasn't figured out how to make money, but that hasn't stopped Web giants Google Inc. (GOOG), Yahoo Inc. (YHOO) and Microsoft Corp. (MSFT) from racing to establish real-time search capabilities.
Real-time search helps Internet users find Web posts, including those from San Francisco's Twitter Inc., seconds after publication. The field has grown in importance amid the exploding popularity of services like Twitter, which lets users blast short messages rapid fire from computers and mobile phones.
The growth of Twitter has fueled expectations that real-time search could drive Internet advertising to new heights by allowing marketers to target relevant ads at consumers interested in breaking events, hot topics or their favorite celebrities. Some proponents argue real-time data and search could develop into a billion-dollar market.
"Every conceivable advertiser will be interested," said Ron Conway of SV Angel LLC, an early investor in Google and Twitter. "It will create a huge monetization opportunity."
Just how that opportunity will unfold remains unclear. There is no shortage of real-time search startups - such as OneRiot LLC and Scoopler Inc., not to mention Twitter itself - that are attempting to make sense of the growing universe of real-time user-generated data. It is telling, however, that even Twitter still hasn't said how it hopes to turn user updates, known as "tweets," into revenue.
Still, Google, Yahoo and Microsoft are pouring time and resources into the real-time Web. All three have had discussions with Twitter seeking some sort of search or advertising deal, according to people familiar with the situation. They also are looking beyond the micro-blogging leader.
The search giants note other sources of user-generated real-time data, such as Web recommendation engine Digg Inc. or micro-blogging services like Tumblr Inc. They also point to their own properties. Microsoft, for example, notes its Messenger and Spaces services are real-time data sources, while Yahoo highlights its Answers service and its experimental Brazilian micro-blogging property Meme. In January, Google pulled back from its Jaiku service, but recent blog rumors suggest it is poised to launch a service that indexes and ranks content from microblogging services, like Twitter.
Making sense of real-time data poses technological challenges for the big search companies. Their current algorithms return results heavily weighed towards older Web pages that have established credibility and attracted large audiences, an approach at odds with real-time search.
Twitter is like a fire hose spewing out a flood of tweets, many of which are seconds old and from obscure users with little track record. Tweets many times contain acronyms, Web site address abbreviations and emoticons, all of which make it difficult for traditional search engines to evaluate their relevance - and filter out "tweet spam."
"Whoever figures out how to filter out spam best will win the real-time search battle," said Kevin Lee, chief executive of search engine marketer Didit.com LLC.
Prabhakar Raghavan, who runs Yahoo's search strategy, says the company is looking at how it might data mine tweets and other real-time feeds, a process that will help it evaluate and summarize content more efficiently. Yahoo is also looking at whether it might map tweets, allowing advertisers to target geographies where interest in a product or service is growing.
Microsoft senior program manager Andy Oakley says his company is also determining how to filter, summarize and present real-time tweets. He suggests up-to-the minute micro-posts and links could be displayed in an "updates" section within a traditional search results page.
Google last month introduced a "recent results" option to its search engine, and co-founder Larry Page has spoken publicly about the need to continually quicken the pace at which the company's spiders index Web pages. A company spokesman said Google was looking at ways to make real-time data more useful to its users.
Tobias Peggs, general manager at OneRiot, says real-time searchers tend to search the Web many times a day because they expect results will be updated more quickly than on established engines like Google or Yahoo. He believes that gives companies like OneRiot more opportunities to serve up relevant ads based on the changing situations.
"If the latest update on Britney Spears says she wore green Gucci dress last night, that would be an opportunity for Gucci to advertise that green dress," Peggs said.
via http://online.wsj.com/article/BT-CO-20090615-712397.html
1 comments:
Post a Comment