At eTail last week Chris Anderson spoke on the long tail. Chris is the editor in chief of Wired magazine and author of the book The Long Tail. He described how the most popular things are way more popular than the less popular ones and when you chart them you get the power law graph that has the long tail that Chris has made famous. I’ve talked about how site search has a long tail in the past and enjoyed seeing Chris in person.

Traditional retailers focus on the products at the head of this graph – the popular products. Shelf space is relatively expensive and the products on the shelf have to sell at a certain rate in order to justify being there. This forces traditional retailers to carry only a limited choice.

Internet retailers are not constrained by this. Because their shelf space is almost free they can carry a much larger inventory. All these products in the long tail are not big sellers by themselves but when you add them all up you can have more sales than you do for the popular products. The really interesting thing from a retailers point of view is that products in the tail have higher margins and higher satisfaction. There is higher satisfaction because people are more likely to find things that match their unique tastes in the long tail and they’re willing to pay for it. This seems to be a win for everyone – by selling products in the long tail you make your customers happier and you make more money.

One of the challenges for internet retailers is trying to work out how to drive people into the long tail. Retailers like Amazon and Netflix are doing this through recommendation engines, (people who bought this book also bought this book). This got me thinking about how SLI, as a site search vendor can help our retail customers encourage their visitors into the long tail. Obviously site search is a window into the long tail. One of the features of our search is our Related Searches. I suspect these do exactly this.

For example if you search on Deep Discount for The Departed one of the related searches is the director Martin Scorsese. If you click on that link then we suggest Mean Streets. So within two clicks you’ve gone from a current blockbuster to a relatively obscure 1973 movie.

We probably need to do some more research but it looks to me like our Related Searches do a nice job of exposing the long tail.

I noticed that Google are experimenting more with related search and someone had spotted an ad where MSN was touting their related search (sorry, I can’t find the link to this). This is good for users and for the search engine’s bank balance.

It’s good for the users because related search is useful and really easy to use. They help people form longer, more refined queries and the net result is that people are more likely to find what they are looking for. I’m pleased to see that Google’s related searches are not necessarily lexically related. For example they suggest crafts for hobbies. This is in contrast to Yahoo’s related searches (labelled “Also Try”) – which are lame because they always contain the original search term. Google’s related searches look less lame than Yahoo’s. I’m not one of the random users that is seeing them – so I haven’t had a chance to test them out yet.

It’s good to show the related searches at the bottom of the page. If a user reaches the bottom of the page then the search results haven’t satisfied them and the related searches provide an easy way to try and find some more relevant results. We do this on all of our site searches. However related searches are so useful they should also be somewhere at the top of the page. Either above the search results (like Yahoo) or at the side (like Ask). I think right at the top is best.

Related searches will help Google’s bank balance because related searches help people find what they are looking for – be it organic or paid results. Search Engines who use our related search have reported 10-15% jump in revenue from paid search. This is one of those nice situations where you improve the user experience and it helps the bottom line.