Long-Tail Criticisms – Applications for SEO

The phrase The Long Tail was popularized by Chris Anderson in Wired magazine in 2004. The statistical model with the cool name is primarily used to describe marketing models seen primarily in large firms like Amazon.com, Ebay, or Netflix.

Rather than discussing the mathematical nuances of long-tail, this article aims at debunking some myths associated with search engine optimization and the supposed ‘holy grail’ status of long-tail. Many SEOs have praised the new information and tout its ‘powers’ as crucial to any web sites success. DistinctSEO will invite a special guest to talk about the ‘pros’ of long-tail, but in the meantime, let’s look practical applications from an SEO standpoint.


The long-tail is the colloquial name for a long-known feature of statistical distributions (Zipf, Power laws, Pareto distributions and/or general Lévy distributions ). The feature is also known as “heavy tails”, “power-law tails” or “Pareto tails”. What does all this mean? Nothing if you can’t understand it. Here is an image to explain the basic premise of the ‘long-tail’ concept:

Long-Tail

The Y-axis represents hits to your site and the X-axis represents the keywords used for each hit. The idea: the red area representing the top searches in your niche area and generally most competitive will equate to the summation of the green area, the less popular and less competitive keywords.

Many SEOs and marketers alike are starting to stress the importance of the long-tail, not only for marketing purposes, but for optimization as well.

“Don’t focus heavily on your the most competitive keywords when you can garner the same traffic by optimizing for the words in the long-tail.”

Although on the surface this appears to be a logical tactic, however, you must go deeper before accepting the long-tail as a crucial component of your SEO objectives and strategies. Let’s look at some issues one by one.

1. Is Long-Tail Practical For All Websites?

Although every web site has unrecognized value hidden in long-tail, is it really worth the time and effort to expand a small web site to reflect the long-tail? Here’s an example: Assuming the top ten keywords for you particular industry account for 5% of the long-tail; you’ve optimized 5 pages for the ten keywords. In order to recognize the remaining long-tail keywords you must build content to ‘capture’ the extra 95%. If 5 pages acoutned for 5% of the long-tail, you need to create another 95 pages to accommodate the remainder. Yes, there are more efficient ways to write content so you don’t have the keyword/web page ratio I’ve displayed, however, for many small web site owners this is a huge investment to consider.

(As an aside, I won’t make the argument that small web sites will run out of content to accommodate the extra 95%. The more distinct and narrow your industry the less keywords you will likely have in your entire long-tail distribution.)

The biggest success stories of long-tail are HUGE online firms like Amazon.com. (For the record I’m guessing their Long-Tail looks more like rectangles than curves….) However, Amazon.com also likely has the largest online inventory; it has been said that every day their long-tail changes completely (yesterday’s top 10 keywords are not today’s top ten.) Can small firms necessarily expand to accommodate fringe keywords? Although the investment of expansion (of your business and the subsequent SEO) may pay off in untapped market areas, one should have the practical business sense to limit their growth to their expertise/distinct capabilties. This may involve ignoring long-tail results. Remember, is your objective to make more sales or to get as much traffic as possible? I’m guessing the former.

2. Does Long-Tail Really Mean More Traffic?

Here’s the thing, long-tail graphs only depict the organic traffic you’re already ranking for. If you concentrate your energies on the long-tail you’re suggesting that a) there is an existing market for those obscure keywords, and b) you don’t already rank in the top five for the keywords (risking spending time where you don’t have to.) Hopefully the LT and other statistical tools will help your draw the line between worthy and not so worthy content.

(There are PPC (pay-per-click) componenets to this question, however, we’ll let our guest writer explain this position in better detail in another article.)

There still, however, remains a dilemma to the SEO/Webmaster. Should you expand content or not? Luckily there ARE practical methods of attacking long-tail keywords. An easy tactic to accomplish this is running an online forum. Each new post and forum category is another opportunity to find and tap keywords you haven’t directly optimized for. Your long-tail may grow since the forum incorporates many categories giving you added insight into your market. Conversely, the long-tail is larger with a forum because you have opted to include more and more content. Don’t get confused on how you derived your LT results.

3. What About the Interval?

Here is a statistical item to consider, what about intervals? The interval is represented by the dashed vertical line in the image above. If you’re a stats guru you probably run your own long-tail data through you own analyses. For those of you who use online programs there are issues to consider.

One Long-Tail program (online) I use ONLY calculates the top TEN keywords as part of the ‘hot-zone’ and let’s everything else fall into the ‘long-tail’. Usually this means my interval sits around 80%. This for some may seem to big or too small. The position of your interval will dictate how big your long-tail is, it can be moved.

By merely considering the top 10 keywords as your ‘hot words’ you could be spending too much time on words that are too competitive. Conversely, if you go too far down the long-tail, all the way to the end, you will spend too much time on words that will never garner returns. To get a better idea where to draw the line one should choose to run their own statistical confidence intervals and standard distributions to narrow the long-tail field. If you’re going to invest the time to expand content to accommodate LT then you mind as well spend the time excluding keywords that may never be profitable.

In conclusion, the Long-tail applications for SEO are viable. This is another tool to be used in trying to decipher how the market reacts and where the traffic lies within your particular niche. However, reliance on supposed SEO wonders that LT will produce is wishful thinking. This is a great content building tool for the a web site ready to accommodate an increase in content/product. Before major SEO and content decisions consider the size of your expansion and what you can handle, and various methods you can accommodate portions of the LT (not the entire thing.) Understanding its applications can certainly expand your traffic, LT can help–converting traffic is another matter.

[tags]longtail, long tail[/tags]