What Keyword Term Lengths Tell us about the Long Tail of Search

A few weeks back I wrote about thickening up your site and the long tail of search, stating that the majority of most site’s organic keyword traffic is actually stored up in the long tail (keyword phrases > 3 words)  and it’s important not to overlook that idea when targeting organic traffic – and so this week I decided to prove the concept true by looking at some real website data but taking things one step further than the traditional long tail theory.

If you were to ask an SEO professional what the long tail of search looks like, you might get something like the below chart:

the broad, competitive terms get the most traffic and the longer, less competitive phrases get less traffic per keyword but still make up a significant portion of the total keyword traffic.  But is the general long tail theory really that relevant for most websites who are looking to improve their organic search traffic or is there something else that might be even more interesting and useful to consider?  

For this case study

  • I took two client websites that we currently service
  • grabbed each site’s respective analytics data over the past month (August 26th to September 25th)
  • did some excel number crunching 

and came up with the below two charts.  The first website had recorded about 2,000 organic keyword searches while the second website recorded about 10,500 (over the course of a single month)

Below is the chart for the first website showing keyword phrase length vs volume.  Notice the smooth looking wave, which peaks at ~500 keyword phrases of length 4 terms per phrase. 

Looking at other long tail studies, I already know broad keyword terms appear to receive the most traffic.  However, the really important point to note here is that we’ve got ~70% of total keyword traffic going to phrases 4 terms or greater, with the remainder going to one to three term phrases.

Pretty cool you say?  Now take a look at the second chart below, showing the second website with about 10,500 recorded keyword phrases.

Yes, both charts are strikingly similar.  I actually didn’t think I’d see such a clean resemblance – two different websites, two different traffic levels, two different sets of keywords, yet we’re seeing the same trend that the long tail (greater than 3 terms) makes up roughly 70% of organic keyword traffic received.

Keyword Term Length Distribution is a Wave and has a Sweet Spot

Note that the excel generated charts above depict more of a wave shape peaking at ~4 words vs what you might see from a typical "SEO" long tail chart (refer to the first image in this post).

So really, it’s more accurate to say that the first one or two head term lengths a website receives doesn’t really have the most volume, as the traditional long tail chart depicts – it’s actually around 3-5 words per keyword phrase that make up the sweet spot in terms of highest volume per website.  Of course results may vary depending on the type of website and size, but I’m betting on average you’d get close to the results I came up with.

Don’t confuse these results with similar findings

My results shouldn’t be confused with studies such as Hitwise’s longtail results, which show the top search terms (not term lengths) by volume from the engines (and not keyword traffic received by a website) – no, the results in this post are depicting the distribution of keyword term length for individual websites, not keyword volume occurrence over the entire internet.  It’s easy to get confused since both are relevant pieces of information and both are correlated with the term "long tail".

Why does long tail traffic matter?

So I’ve told you that long tail traffic makes up the majority of organic keyword traffic that a website will receive, on average.  If that weren’t important enough…

Imagine that you are searching for something specific on the web.  Maybe you’re trying to find a new cell phone battery or the best place to eat sushi in your local neighborhood.  Would you type in one or two word phrases to find what you’re looking for?  As it turns out, no, you’re more likely to use longer phrases to more accurately describe what it is you’re really seeking out – the above data charts reinforce this concept.

And people who are specifically looking for something are more likely to convert – now that’s gold!  And if you’re interested, Best Rank offers content writing services to compliment any/all online marketing campaigns to help tap into long tail traffic.

Want to repeat my experiment?

  • Just open up excel
  • paste all your keyword phrases into a complete list.  
  • Next create a new column representing the number of words in each phrase.  To do this, you can use the following formula, just copy and paste down the entire column:

=IF(LEN(A2)=0,0,LEN(TRIM(A2))-LEN(SUBSTITUTE(TRIM(A2)," ",""))+1)

where "A2" is your cell containing the keyword phrase.  

  • Then you will want to count all occurrences of each number (1 through N, N being however high you think the longest term in your results might be).  Use the following formula in each cell, 1 through N:

=COUNTIF(range,"text")

where "range" is something like "A2:A5000" (cells A2 through A5000) and "text" is your integer, "1" for example.  

  • Finally, make a chart from the resulting data and you’re likely to get something very similar to my results.

What do you find?

Connect with Mike Shannon on Google+

5 Comments

  1. William Alvarez

    I find it really interesting Mike, and it certainly is true for many websites, but what you didn’t mention is which verticals these two clients fall into and how aggressive they are with their efforts. I would say that if any of them lead the head with broad terms then it would be a different story because they made it to top ten and are benefiting from traffic from high search volume queries, but if they are not top of mind, then there is a notorious opportunity with keywords in the long tail for them which makes more sense for players who come late into this competition or are shy with their tactics. I’m assuming this is what you wanted to point out. Thanks for the formula!

  2. Shane Halstead

    I have similar report, but I’ve added secondary view that removes the website’s branded keywords as they can skew the data. So if I’m working with a company called "Acme Auto Parts", the secondary view removes all attempted iterations of that branded term.

  3. Mike Shannon
    Mike Shannon

    Hi Shane, both brands in this case are two letter words, so branded terms in themselves would come in at 1-2 words.  A long tail phrase that also includes the brand keywords is still a long tail phrase and in this case we’re seeing a peak around ~4 words per query.  Interesting stuff though, I’m sure there are a lot of ways to slice the data and that the surface is just being scratched.

  4. Mike Shannon
    Mike Shannon

    The two sites are in two different verticals, and since both sites are a client of ours we’re at least trying to be somewhat aggressive ;)

    You make an interesting point, in that if a website is right at the top of the rankings for a high traffic keyword then the keyword chart might look more skewed toward the broad end… maybe that is true, however I’m assuming that a search engine won’t push a website to that point until there is enough trust: enough pages in the index, links and enough long tail queries sort of "supporting" the broad terms.

  5. margaretmetchell

    Thanks for the post! I’m glad that came across it because I found here a lot of interesting information for me. I want to add that as a search agency we see many different styles of PPC campaign design, from in-house teams, other agencies, and occasionally running across those from the search engines themselves. Broadly speaking, campaigns can fit into one of two categories: 1. Those heavy on keywords or phrases made up of low number of fairly generic terms (for the purpose of this article, referred to as “shorter keywords”) 2. Those using a high number of terms ranging from generic to improbable (or “longer keywords”)

Leave a Reply

Your email address will not be published. Required fields are marked *

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>