The “eighty-twenty” rule, or Pareto’s principle, is bandied about when trying to express the distribution of many real-world variables, from share of national wealth to causes of industrial production failures. The rule derives from an observation of Italian economist Vilfredo Pareto that 80% of the land in 19th century Italy was owned by 20% of the population (see Power Law for a more general theory).
In the field of Search Engines (or decision engines if you prefer), the Pareto principle does not hold. Eighty percent (some would say ninety) of Internet searches carried out in the United Kingdom are serviced by a single company – Google.
Microsoft have decided to take some market share from Google, and have launched their newly rebranded search engine – Bing. My first thought – really?
Is Bing any good?
Microsoft are not exactly newcomers to the search engine game, having started over ten years ago with MSN Search. Nor are they the only team that has tried and failed to better Google. Thus while I have no ill will to the guys at Bing, I suspect their latest effort is at best quixotic; at worst a diversion of effort from other more productive activities.
Having said that, Bing does have some nice feature. Or at least it does in the North American version; the UK version is still in beta but should offer the full suite of functions by the end of 2009.
If you are searching for a commercial commodity – a hotel room, a seat on a place, a specific model of camera – Bing (North America) can show you price trends to give an indication of whether your proposed purchase is likely to get more expensive. And search results have a category drill-down bar on the left with a bit of clever context sensitivity for increasing the specificity of your search query.
For more general information searches, Bing’s (North American) interface has enhancements such as video thumbnails auto-playing on mouse over that are good fun, and the relevance of results is comparable to Google for simple queries, but not as finely grained for more complex matters.
For example, have a look at a side-by-side comparison of two queries using Google and Bing:
- Query: “Michael Jackson” (North America)
- Query: “What is the median household income in Bristol, UK?” (UK)
Results are comparable for the first query, and Bing’s emphasis on video and image results is attractive. But for the second query, the relevance of Google’s results is clearly much stronger.
Is Bing going to take searches from Google?
Probably not. Or maybe. Or actually I really have no idea. I mainly write about local politics; what the hell do I know?
There are two problems I see about trying to be better at Googling than Google.
The first is that the value that a typical search user gives to their result is heavily biased by the search provider. In a study conducted by a team at Indiana University [authored by a guy called Bing] “In Google We Trust: Users' Decisions on Rank, Position, and Relevance”, the results suggested an implicit trust in the Google brand. The impression that “Google tends to returns the most relevant results” transforms into “Whatever Google returns is the most relevant result”. Thus even if you start a new search engine that is in absolute terms better than Google, the very fact that the results are slightly different from Google may count against you.
The second problem with innovation in search is that the dominant presentation of a search is a screen with a box into which one types text and then receives a list of links. This model works and matches the expectation of the consumer. There isn’t that much opportunity to experiment with the front end without scaring people off, which is probably why Microsoft’s approach to differentiating their product is to put a picture on the search screen and spend $80 million on marketing (On that note, thanks for the beers, guys). But even Google is constrained by customer preference - their exotic search interfaces like Timeline and Wonder Wheel remain fairly low-key.
Video: Stefan Weitz discusses Bing
Summary
So – Bing. The North American version has some compelling features for searches relating to commercial transactions such as “How Much is…” or “Where can I buy a…” and information requests of the form “Who is…” or “What is…”.
As yet, Google is still a better starting point for more abstract queries such as “Tell me about…” or “Is X better than Y?”. But most searches are not of this character. If you’re a UK reader then as soon as all the clever functions go live, Bing is probably worth a go.
Personally, I wouldn’t spend hundreds of millions of dollars trying to get back into the search market for the developed world. But 75% of the world population are not Internet users. So perhaps if another 10-20% come online in the next decade, Bing might be able to offer them something compelling, and they might even change our expectations of what a Search Engine could be.
Related Links
Other Bloggers/Tweeters/Liggers at the Microsoft event:
Phil Bradley - http://www.philbradley.typepad.com/
Tom Warren - http://www.neowin.net
Chris Hambly - http://www.chrishambly.com/
Josh Russell - http://www.joshrussell.com
David Stuart - http://blog.webometrics.org.uk/
Karen Blakeman - http://www.rba.co.uk/wordpress/
Neil Robinson - http://www.blog.lanzen.co.uk/
Kevin Harrington - http://www.kevinharrington.com/
Mark Sharp - http://www.perspicuousasmud.com/
Phil Leggetter - http://www.leggetter.co.uk/
Andrea Giammarchi - http://webreflection.blogspot.com/
Bookmark/Search this post with: