Does putting two losers together make a winner, or just one big loser?
Microsoft CEO Steve Ballmer has declared that combining his company with Yahoo would make the internet marketplace more competitive by establishing a strong number two in search and search advertising.
Then again it, maybe it will just make a weak number two. Microsoft and Yahoo have both been leaders in their own fields but in search, they suck. Put them together and they will still suck, in a bigger way.
These days it’s search that matters and the customers are discerning about it. The key to success is not brand or scale but the quality of the algorithm, and Google’s is the best.
Maybe Microsoft and Yahoo together can invent a better search algorithm than Google, but it’s doubtful. They haven’t done it separately and anyway Google is now making so much money it is attracting the best programmers who are constantly refining the algorithm that Sergey Brin and Larry Page invented in 1995.
It was then called ‘BackRub’, named for the way the algorithm analysed the “back links” that point to a website. Now it’s called PageRank, after Larry Page, and although the algorithm has been refined and added to over the past ten years, at its heart the system still computes a recursive score for web pages based on the weighted sum of the other websites that link to them.
That was the insight of Brin and Page: that the number of links to a website that are generated by other human beings is a good way to measure relevance based on human concepts of importance.
The other advantage of Google was that it focused on search. That meant it has a beautifully simple website, with nothing but a good brand, a search field and – for the first few years at least – the number of web pages being indexed. That number passed a billion pages in April 2000, which made the Google search offering astounding and compelling.
In February 1994, a year before Brin and Page started creating their search algorithm, Yahoo’s founders, Jerry Yang and David Filo, working in a trailer at Stanford University California, started organising their own web surfing into links that they saved as bookmarks. As the list of links got longer, they put them into categories and then sub-categories.
It was a pretty useful bunch of links, so they created a web page to make them easier to use and available to others. It was called ‘Jerry and David’s Guide to the World Wide Web’. That soon became too long so they used an acronym: Yet Another Hierarchical Officious Oracle – that is, Yahoo
Word spread fast around the internet and Yahoo had its first million hit day in the autumn of 1994, which translated to 100,000 unique users. In April 1995 Yang and Filo got $2 million from Sequoia Capital, which had seed-funded Apple Computers, Oracle and Cisco Systems, and they were away.
So Yahoo is a directory site that added search later, when the company bought the Inktomi search engine in 2002.
Inktomi was developed in 1995 by Eric Brewer, at the University of Berkeley, and put more emphasis on keyword density than external links. It was a sort of wholesale search, designed to power the search engines of other websites, including AOL. In 2003 Yahoo also bought Overture, another wholesale search business.
Microsoft, meanwhile, was sound asleep, lying on the soft bed of the cash it was making from its Windows operating system monopoly. It eventually used the Inktomi/Yahoo search engine in msn.com, but basically Microsoft was caught out by Google’s advance in the same way IBM was caught out by Paul Allen’s and Bill Gates’ PC revolution during the 1980s.
The only other major search engine that survived the shake-out that followed the tech wreck is Ask Jeeves, later renamed ask.com, developed by Garrett Greuner and David Warthen. They had the idea of users being able to search using plain English questions; it was pretty cool, but never really took off, although Ask is still one of the top four search engines.
So there are four major search engines, but it’s Google, daylight, and then the other three.
The reason Google dominates is that its algorithm is still the most efficient – it most quickly searches the most pages and sorts them better than the others. It’s true that its brand is powerful and the advertising model produces so much cash that the company is a juggernaut, but those things derive from the quality of the search engine.
Google now is the market. Search engine optimisation (SEO) means manipulating or subverting Google to get a better ranking, and Google spends a fortune trying to maintain the integrity of its results. For example, there’s a ‘sandbox’ that you sit in for a few months while Google’s web crawlers inspect your website to make sure it’s kosher.
Brin, Page, Yang and Filo have themselves proved that on the internet scale does not matter – they succeeded when their businesses were small.
One day someone will challenge Google, but it’s unlikely to be Steve Ballmer.