Microsoft's Bing needs more than advertising to compete with Google

London Microsoft's new search engine, Bing, went live on June 1, though the UK version is in beta and lacks many of the features in the US edition. Although the core index is the same, Bing is not just a rebranding of Microsoft Live Search, but has new features including instant previews of websites and videos, automatic categorisation of search results, and Best Match results with deep links.

What about the actual results, though, are they as good as those from Google? Michael Kordahi, a Microsoft employee, reckoned that testing search could be like tasting wine, where you cannot help being influenced by the label, so he created a blind search engine. Search results are returned in three columns with identical formatting, representing hits from Bing, Google and Yahoo! (the order varies), and users vote for the best one. His site went live on June 7. Initial results slightly favoured Bing.

A day later, Google had pulled ahead. "Google: 45 per cent, Bing: 33 per cent, Yahoo: 21 per cent of 8,518 votes" reported Google's search associate Matt Cutts in a Twitter post. Shortly afterwards, Yahoo's vote soared, but by then it was obvious something was wrong. "[Someone] is gaming the system, I've removed the ability to see the results until I sort this out," reported Kordahi.

The result is that we are no closer to knowing which search engine generates the most favoured results, though early indications suggest while Google may be slightly preferred, its margin of success is less than its market share. According to figures from Net Applications, Google gets around 82 per cent of searches, against 9.5 per cent for Yahoo and 5.5 per cent for Microsoft.

That raises the question: Is there anything Microsoft can do to wean us away from Google?

The search giant is the default in Apple's Safari and Mozilla's Firefox, and it is common for new PCs to come with Google's search tools pre-installed. According to Advertising Age, Microsoft is budgeting $80 million to $100 million (Dh294 million to Dh367 million) to promote Bing, but it will take more than advertising to change people's search patterns. The evidence so far is that Bing saw an early spike in usage, driven by tech-savvy users giving it a try, only to sink back a few days later as old habits returned. Bing needs to be dramatically better than Google.

Being almost as good will not win new users; though even Kordahi's blind test is not perfect, as it hides the new usability features that Microsoft is promoting. The consolation for Microsoft is that most reviewers have found Bing better than expected, and that while Google will be untroubled, winning a few points of market share looks possible, though by no means assured.

There is still potential for new approaches to search though, and to prove it the best tool for researching what happened to Kordahi's search experiment is neither Google nor Bing but rather Twitter.