Search’s Long Tail

The Long Tail,” by Wired editor Chris Anderson, made an enormous impact last October. It discusses how the media and entertainment industries will succeed not by only pushing popular, mass-market hits but by also mining the “long tail” of interest among a few people in less-popular media and entertainment properties.

Sure, thousands of people may want to buy a hit song. But add up all those who want to buy lesser-known titles, and they could generate as much, or more, revenue than the hits. As a merchant, you want to tap into both the head of interest and the long tail that trails behind.

The tail makes more sense when you see it illustrated:


Those are the top 100 queries related to shoes on Yahoo’s network, from an article I wrote last September about Yahoo Search Marketing shifting to a broad match system.

A large number of queries happen far less often than leading terms, such as “shoes” or “running shoes,” which are at the “head” of the list. Most queries form the long tail illustrated behind the head. Tap into that tail, and you’ve got substantial traffic — traffic that often converts better than less-general terms.

Search has a long tail, too. I’ve long heard references to a “search tail” or “query tail.” Tails aren’t a new concept to search marketers.

It’s great the tail concept is becoming more popular. It’ll help search marketers who mistakenly fixate on only the most popular terms realize they must also consider the tail.

Over the years, I’ve referred to the search tail as getting the “onesies and twosies,” those queries that may only occur once or twice in a month. Perhaps they don’t appear important because of their individual low volume, but tap into lots of onesies and twosies, and you can do well.

If you haven’t been mining the long tail, start now! Consider broad matching on both Google and Yahoo For organic SEO (define), create lots of good content that naturally taps into a query’s tail.

Further Reading

Further reading illustrates how the search tail is becoming popular and integrated as part of the more general long tail awareness:

  • The long tail of software. Millions of Markets of Dozens“: From the blog of Joe Kraus, one of Excite’s cofounders, this contains a chart of the Excite search tail from Kraus’ day. Kraus says 97 percent of the service’s traffic came from the tail and the real reason his company went out of business was because it didn’t know how to make money from that tail.

    I’d beg to differ. Excite gained search-targeted listings through a partnership with in April 2001. Over six months later, Yahoo finally obtained the same through its partnership with Yahoo Both partnerships let these companies earn from the search tail. Excite was ahead of the game, yet died. Yahoo survived. Lack of tail targeting? No, something else was at work.

    Excite went bankrupt because, among other things, it couldn’t afford to continue competing in the expensive portal competition. Maybe if it had gone after the tail even earlier, it would have survived (though this was hard).

    Yahoo pulled through, so failing to target the tail can’t take all the blame. Also, though Excite the company died, Excite the Web site is still alive. Traffic is much reduced, but today it earns revenue for InfoSpace, the company that runs search there.

  • Google’s Long Tail“: On Chris Anderson’s Long Tail blog, you can see an example of how Google tried to tap into the new understanding and popularity of the long tail concept last February, to illustrate to investors how it’s a long-tail oriented of company. Google’s always been this way, but people now better understand what it does (and what Yahoo Search Marketing pioneered back in its GoTo days).
  • On the Trail of the Long Tail“: Also from February, this post on the Yahoo Search Blog has the company claiming to be a long-tail company. Sadly, it didn’t provide charts to help illustrate this.
  • The Paid Inclusion Dinosaur“: My column from last June looks at why Yahoo has stuck by its guns to run the paid inclusion program. Doing so lets it tap into the giant tail of queries even marketers using broad matching may never consider, especially when you understand paid inclusion ties into editorial results that are far more likely to be clicked on than ad listings. From the longer version for Search Engine Watch members:

    Right now, Yahoo and Google struggle with how to fill out the “tail” of search queries with ads. These are queries that happen only a few times per month but in aggregate represent a huge amount of unsold inventory.

    Advertisers focus on high-frequency terms such as “shoes,” for example. Yahoo reports the term had over 1 million requests in April 2004. It sells for a top bid of around $0.55, with 90 advertisers competing for it.

    In contrast, “winged track shoes” was queried only 27 times last month. No one’s bidding on the term. That’s most likely because the low frequency doesn’t make it seem worth the effort. Yet someone searching for something so specific might convert better than a more generic “shoes” searcher.

    Paid inclusion is a perfect solution for unsold ad inventory like this. Simply tell the advertiser you’ll spider his site and let pages appear in the paid placement area for a flat, low-cost rate when space is available. Advertisers can still target the important terms, but they gain easy visibility for other important ones. Yahoo reduces its unsold inventory. Searchers are spared confusion.

  • Search Engine Marketing Goes Mainstream“: This report from our Search Engine Strategies Boston conference in March 2003 covers how search’s tail was discussed in economic terms and as a search marketing tactic.
  • Just Right: Targeting the Tail“: This is a cached copy of a LookSmart presentation at SES Australia in March 2003 that says all the things you’re hearing about today regarding the long tail. It mentions the “keyword tail” and the “search tail” and concludes:
    • The top 50 percent of queries equal 80 percent of volume.
    • The 80/20 rule doesn’t regularly apply to search.
    • Focusing on just the first 20 percent of keywords will invariably miss most of the value.

Want more search information? ClickZ SEM Archives contain all our search columns, organized by topic.

Related reading

Brand Top Level Domains