A version of this piece was published in Marketing in 2008
The story of David and Goliath has attracted people over the generations, for its tale of hope for the small guy, for triumph in adversity, overcoming the odds. Hannibal, the Cartheginian commander, is remembered not just for that terrible film with Oliver Reed, but amongst military tacticians for his victory at Cannae where he destroyed a Roman army which massively outnumbered him.
This celebration of the underdog reflects our need for heroes, but also our desire to control and check the powerful.
So the launch last week of Cuil (pronounced ‘cool’) a new search engine from some ex-Google staff is interesting not just because of what it is, but what it says about the incumbent.
What it is, is impressive. Cuil claims to index three times as many sites as Google, using an algorithm that looks not at popularity (the number and quality of links pointing to a site) but at context – examining the context in which the searcher’s keywords sit in a page in order to understand better the relevance to a search.
If you look at search in purely rational terms, the user is only looking in reality for one result, not thousands. But Google’s claim to index billions of pages, and then its practice of presenting thousands of them to the searcher, give the illusion of breadth and choice – regardless of whether they find the thing they’re looking for. This is a powerful proposition, and it’s one the Cuil clearly have in mind when they promote themselves as listing three times as many pages as Google.
There’s another noteworthy feature of Cuil – their attitude to privacy. Coming hot on the heels with Google’s experience with a US court demanding they hand over all the viewing data for every user on YouTube, privacy is a weak spot in Google’s armour. Google hang on to the search data of every user for eighteen months – a fact that unsettles privacy campaigners, particularly after AOL’s disastrous public release of thousands of searchers’ data two years ago.
Cuil keeps no data from its users’ searches, and it makes a point of this. They keep no log files, IP addresses or personally identifiable information – as they put it, “Your search history is your business, not ours”.
The search results themselves seem promising – an interesting and different layout to Google’s, with three columns of text and pictures. A sliding box to the right allows the searcher to drill into different categories – so a search for Orange brings up the mobile phone company as first listing, but categories allow you to focus your search into Orange County, California, Orange Sodas or Citrus fruit.
This is smart and useful, because the engine is making a reasonable fist of differentiating between fruit and phones – for which it needs to understand context. It doesn’t always get this right – Sky TV comes first in the listings for a search for ‘sky’, but there’s a Virgin Media logo next to the listing.
So Cuil’s interface is a useful improvement on Google’s. But good search and a useful interface isn’t enough.
A quick look at Google’s Q2 results (which despite disappointment in the markets were still 39% up on the previous year) shows that 31% of their revenues come from partner sites – other websites which carry the Google search box.
Distribution is the key to Google’s massive success – most sites take Google’s AdSense program, and Google pay big money to them; $1.47billion out of the $1.66bn they earned that quarter.
So Google’s success in gaining distribution is down to two key things. The monetisation they drive from each search is better than their competitors, and their ability to model and predict this is better – so they can cut better deals for distribution partners, with stronger guarantees.
This is the barrier to entry for Cuil. It isn’t just the quality of their product – it’s their ability to drive distribution that will determine ultimate success. So Cuil is an interesting launch, but the chances of it launching that single stone that strikes the giant between the eyes are pretty slim.