A guest post on the Hitwise blog by search engine optimizer / analyst Dustin Woodward, suggests that the long tail of search might be much longer than most of us expected.

Woodard evaluated a data set of search terms used in the last three months, and found that the top 1,000 search terms only accounted for about 10 percent of all search traffic.

Indeed, the Hitwise’s top 100 search terms account for just 5.7 percent of all search traffic, while the top 1,000 search terms only account for 10.6% of all search traffic. Even the top 10,000 search terms drive just 18.5 percent of all search traffic.

What does this mean for Internet marketers? Well, it does NOT mean you should try to optimize your site for all those long tail keywords. Together they may account for a lot of searches, but if anything, Woodard’s analysis confirms the fact that long tail terms are infrequently searched on i.e. they are likely to be ‘one-offs’.

There may still be opportunities to pick up traffic for such one-offs, but trying to target such keywords is unlikely to pay off – simply because such words are unlikely to be searched on next month.

So it still pays to optimize for the most popular, relevant keywords you can afford to target. And if you want to be creative, think into the future rather than the past. In other words, instead of targeting last month’s long tail keywords, think about what keywords you think will become popular, based on trends in your market or industry.

Source: Frederic Lardinois, “Hitwise: The Long Tail of Search is Much Longer Than Expected”, ReadWriteWeb, November 7, 2008

You may think you know the answer… but you may be surprised…

Magazines.com recently ran several tests that, among other things, compared the results of using the homepage as a search landing page compared with using a product-specific landing page.

The Magazines.com team, led by Kate O’Neill, director of customer experience and product development, ran a two-week test to see how the Magazines.com homepage - featuring a highlighted special offer on O magazine (a 67% discount for a 1-year subscription) would fare against their O magazine product-detail landing page, for search traffic.

Ms O’Neill then expanded the test to include other magazine titles, with markedly different audiences to those of O magazine. In each case, the homepage - including a special offer highlighted for a given magazine - was tested against a title-specific page for search traffic.

Surprisingly enough, O magazine search traffic converted more when they landed on the homepage/special offer combination than when they arrived at the product-detail page. But for the other titles tested, the title-detail page outperformed the homepage/special offer combination.

Once again, these results highlight the fact that different audiences respond differently to the same elements and… that you need to TEST before reaching any conclusions about what will work best for a given product or target market.

We’ll undoubtedly return to this theme again and again, as tests we’ve run on our own sites and those of our clients continue to bear this conclusion out.

Source: MarketingSherpa, “Homepage Offer vs. Product Landing Page - Search Test Results”, MarketingSherpa, June 19, 2008