Interesting read by Randall Stross from The New York Times re Google and the economics of the Internet search business. Salient points follow -
The most valuable secret formula in American business used to be Coca-Cola’s. Today, it’s Google’s master algorithm. In the search business, however, there’s no rival to play the role of Pepsi. Yahoo is the closest but still a distant No. 2, and Google earns more profits in a single quarter than Yahoo does in a year. Microsoft, an even more distant No. 3 , can’t keep up, despite having USD28B in cash iat the end of March.
The fumbling of Google’s largest challengers, however, has not dampened the enthusiasm of entrepreneurs and venture capitalists for entering the search game. The combination of low start-up costs and potentially huge profit makes it seem a reasonable bet.
Developing a search algorithm can be accomplished by very small teams. It was a team of two — Larry Page and Sergey Brin — who developed a new and improved search algorithm. They beat out Alta Vista, whose search engine was developed by seven people at DEC.
Profit margins in the search business are mind-boggling. Google’s net profit margin last year was 29%. Amazon’s was a mere 1.8%. Even gathering the crumbs left behind by Google could generate a lofty market capitalization. In his blog, Don Dodge, a Microsoft manager argues: “Why 1% of Search Market Share Is Worth Over USD1 Billion.” He reasoned that 1% of the 7.3 billion searches performed in the US alone in March, multiplied by 12 cents in advertising revenue per search, would yield annualized revenue of USD105M. Assuming a market cap that is 10 times revenue, his arithmetic leads to a billion-dollar company.
Another attraction: Any company that offers a superior search service would be able to poach the customers easily. No long-term contracts binds Google users. In addition, even though each search engine conducts a search in its own way, there is no daunting learning curve to prevent users from shifting. Search users merely type in the word or a phrase — no matter which service they are using.
Engines like Hakia, Accoona and Powerset are trying to grab market share by writing a more sophisticated algorithm. A growing number of entrepreneurs are placing their bets, however, on a hybrid system that puts humans back into the search equation. They are grouped under a newly coined rubric, “social search,” and it is becoming a crowded field.
Newcomers like Squidoo, Sproose and NosyJoe offer search results based on submissions or votes by users. Bessed also relies on users to suggest the best Web pages for a topic, but then has editors refine them. ChaCha gives customers the opportunity to have an online chat with a human being who can provide search assistance.
Sometimes a small variation on an existing idea is enough to make it stand out. In Oct 06, when Bessed began its search service with the manually edited results pages, it had only two editors and covered just a few hundred search terms suggested randomly by users.
Last month Mahalo inaugurated a search service with manually edited results. It started with several advantages: VC backing, 30 editors, systematic focus on the most commonly requested search terms, and the added idea of supplying Google’s search results for any search not covered by its own best-of-the-best lists.
Mahalo now has pre-prepared pages for 5K terms related to entertainment, travel, health, technology and other subject areas. It plans to expand its coverage to 10K terms by year-end, and eventually to provide results for one-third of the most common search terms.
The company is financed by Sequoia Capital, an early backer of both Yahoo and Google. Sequoia, like other Silicon Valley venture capital firms, offers experienced entrepreneurs an office and salary to figure out an idea for a new start-up.
It was while he was an entrepreneur in residence that Jason Calacanis,36, thought of Mahalo. He previously published The Silicon Alley Reporter in the mid-90s and went on to be a co-founder of Weblogs, which was sold to AOL in 2005 for about USD25M. He took up residence at Sequoia in Dec 06, founded Mahalo and gathered two rounds of financing, including backing from the News Corporation. He claims to have enough financing to provide five years of experimentation and refinements.
A hand-built Mahalo search-results page has one conspicuous advantage over Google’s: grouping into subthemes, which make a page of links much easier to scan and to find items of particular interest. Mahalo’s page on Paris Hilton, the site’s top search subject last week, lists more than 80 sites, but arranges the recommended links into clusters including news, photos, gossip, satire and humor. The use of subject categories also eliminates the need to provide, as Google does, two-line text excerpts from the listed sites to provide clues about the site’s contents. Thus, they are easier to skim than the 12 sites that fill the entire first page of Google’s search results.
More importantly, all of the links listed in Mahalo send the user to Web pages that contain genuine content, not sales pitches in disguise. By using its own editors as the final arbiters of what goes in, Mahalo cuts off access in its listings to Web sites that confuse a search engine’s algorithm with advertorials that commingle ads with noncommercial information. To those in the trade, outsmarting the algorithm is called “search engine optimization.” For the rest of us, it produces Web pages littered with spam.
Last week, Mr. Calacanis tried to illustrate how spam has infested some top results on Google. After running searches for “low-carb diets,” “Lasik” and “lingerie” at Google and at Mahalo, he compared the results. The exercise succeeded in exposing a few examples of Web sites ranked highly in Google’s results that contained advertorials or content apparently scraped from higher-quality sites.
Google contends that its search engine relies on humans and machines. Matt Cutts, head of Google’s Webspam team, said users who place links on their own Web pages pointing to other sites provide the raw information about valued sites that is incorporated into Google’s PageRank algorithm. But determining the best way to utilize that information requires continuing work by human engineers.
Thus at Google the machine has the final say. Once the query is fed into the “engine,” the results are presented without manual adjustment. At Mahalo and other “human powered” sites, the machine performs a first cut at the search in advance of a user’s request, and the results are then winnowed and shaped by human editors, then stored, a process that Mr. Calacanis terms “editorialized search.”
HUMAN-POWERED search may be able to cover a wide swath of queries if it can draw on the enthusiasm of contributors who have made Wikipedia a phenomenon of huge scale. Mahalo just began a “Mahalo Greenhouse” service that enlists users who are passionate about a particular subject to write a page of search results for USD10 to USD15. Submissions must pass the scrutiny of the site’s full-time editors before posting.
Even Google is interested in exploring “human powered” search. It has combed through its own Web pages to remove all references to “automatic ranking” to prepare for the possibility of relying on user feedback to improve search results or other approaches that are more directly “human powered” than the algorithm.
0 comments:
Post a Comment