There has been a lot of talk in the last year about the 'long tail' in search engine marketing (SEM). What is a 'long tail' and how can it benefit your business?
The term Long tail was coined by Chris Anderson in Wired magazine back in 2004. The premise is while most sales are made for a few 'best sellers', if you have a wide enough variety, the sales in less popular items will surpass those of the 'best sellers'.
Taken to the world of SEM, it means while a large amount of traffic can be generated by successfully targeting a small set of relevant keywords, there is more potential traffic from variants of that set. Estimates are that 20% to 30% of all searches for an item will use a small set of keywords related to the item. The rest will be using variations of that term. These variants maybe synonyms or the keywords with modifiers.
So, does this work and how can you take advantage of it?
Well, yes it does work - sort of. Anderson used the case of Ecast, an online music distributor, with a catalogue of over 150,000 tunes. In this case there is sufficient depth in the 'long tail' to generate the required results. Even if the last 100,000 tunes only had one sale each. The same is true for the online searches. Given a wide enough set, say all searches, and a large enough area, say globally for the last year, the theory would seem to work. However, that is hardly your target audience.
How does it work on a more specific target?
Our website needs an update and I will use it to test the 'Long Tail Theory' as it applies to a small business site. The first page to get a face lift will be our online marketing. Some ideas on possible keyword phrases are 'online marketing', 'web marketing', and 'internet marketing'. I will look at possible other phrases to target in the 'long tail'.
To gather the necessary data, I will use Google Adwords Keyword Tool ( https://adwords.google.com/select/KeywordToolExternal) to get data. While word tracker would have been a better tool I want to use one that is freely available.
To get the data simply enter the first keyword phrase into the box and enter the secret code. In a few seconds you will be presented with list showing search volume for the last month and year. We will repeat these steps for each of the three phrases. The reason is Google only returns 150 or fewer results. By doing them one at a time we can get a larger data set.
The results are targeted to the US market, so I need to them to target Canada. Clicking on 'Edit' in the line "Results are tailored to English, United States", I selected Canada from the dropdown list. (of course if you are targeting the US, you would skip this step.) This resulted in a list of 70 to 150 phrases globally some of which were not found in the subset of Canadaian searches. There was also a list 50 other terms, synonyms, suggested by Google - but I am going to ignore them for now.
Finally, I clicked "Local Search Volume . . ." to sort the list by volume, scrolled down to the bottom of the list and downloaded the csv file. After repeating the process for the other terms I merged the three csv files in my spreadsheet. Giving a list of 370 phrases globally and 265 phrases for Canada around the terms.
The ideal graph for the 'Long Tail Theory' would be smooth and stretch to infinity. Graphing the terms against the data for Canada we get what we would expect, rough graph with 265 data points.
However, what I do have is a list of 265 possible keyword phrases. And, by filtering for Canada, my target market, I have some terms such as "local internet marketing" that seem uniquely Canadian. Of course targeting a single page for more than a three or four related keyword phrases is difficult to impossible. But the list is also useful as a source for topics on blog posts and other articles.
The most searched phrase, not surprisingly was "internet marketing", I decided to use that as the base for my 'long tail' selection. So I will be optimizing my web page for "local internet marketing", "internet marketing service", "bc internet marketing" and "internet business marketing".
In my next article I will report on the success, or failure, of these 'long tail' terms to deliver traffic to the website.