Seasonal Keyword Research (Or: What You’d Wished You’d Remembered from Math Class)

'Tis the season for ecommerce sites to salivate over the visions of sugarplums and higher volume sales dancing in their heads. And in terms of organic rankings, if you haven't prepared your site (or your client's site) to be in optimal rankings in time for the holiday rush, it's too late at this point.

But all hope's not lost--seasonal searches don't only spike during the holiday shopping season. And knowing when searches (that are relevant to your site) see a jump in volume can be key in getting your site in front of the right people at the right time.

Internet marketers—the efficient ones, anyways—spend a lot of time thinking of ways to optimize not only content, but the returns on their efforts. More specifically: we want to see maximal returns while minimizing man hours.

This internal form of optimization is what brought about the notion of capitalizing on the long tail of search queries: you forego quantity for quality (in terms of traffic) while side-stepping the highly competitive short head of search terms. In other words, you can get a page to rank with relative ease while cutting out less qualified traffic, or traffic which isn’t as far along in the purchase cycle.

The long tail looks at keyword search volumes from an absolute or static point of view, but we all know that search volumes fluctuate constantly. However, local search volume or search trends throughout the year tend not get as much attention as global or average search volume numbers do.

While it’s easy to tell that queries involving holiday terms will spike sharply in the weeks leading up to said holiday, you may not be aware of less obvious (though equally seasonal) searches. But if you’re aware of this, you can time your SEO efforts so that you achieve peak results in concert with peak search volumes.

So if you read on, I’ll show you how to tell quickly which keywords in your campaign may have seasonal traffic you’re not aware of. Once you are aware of them, you’ll be in a better position to make sure your client websites are in the right place at the right time when people turn to search engines with seasonal queries.

Get Seasonal Data

The first thing you’ll need to do is get seasonal search volume data from your keyword research tool of choice. I’ll use the Google AdWords tool for my example here since it’s free for all and provides the necessary monthly figures for us to run our calculations.

Of course, with the AdWords tool, you’ll need to select “Search Trends” from the display columns drop down menu. (Optionally, you may also want to select “Highest Volume Occurred In” as well, since that will save you some time after you see which terms are highly seasonal, so you can know which month to aim for rather than having to comb through columns of numbers looking for the same.)

Export your keyword set to excel. For my example, I used “stuffed animals.”

Columns A – P are what the AdWords tool outputs, and columns Q – S are the keys behind the trick of easily calculating what keywords have seasonal traffic. (FYI: the data is sorted by column S, which is the data we’re most interested in.)

If you only want to know the Excel formula you need to drop into your spreadsheet, look no further:

For a data set (for the monthly figures) spanning from column D through column O (see the example above), the formula would read:

=stdev(d2:o2)/average(d2:o2)

If you want to know a bit more about what’s going on with the formula, read on.

What we’re finding with that formula is the coefficient of variation (cv), which tells you how data in a data set is dispersed in relation to the average. Excel doesn’t have a built in function for cv, but as you can see it’s not a complex equation.

Rather, it’s quite simple: to find the coefficient of variation, you simply divide the standard deviation (σ) by the average or the mean (μ). Standard deviation measures how tightly or loosely data is grouped relative to each other, and the average is self-explanatory.

In my example, I calculate both σ and μ in their own columns just to see the data trends, because when I first approached this problem, I thought using the standard deviation would be enough to find seasonal keywords. But as you can see, the most seasonal keyword, “christmas stuffed animals” actually has a lower standard deviation than other keywords listed. However, since it has a lower average value (meaning that search volumes throughout the year are typically very low), that spikes the coefficient of variation. In other words: this keyword is highly seasonal.

What is interesting though is not that “christmas stuffed animals” is a seasonal keyword. (After all, you should be quite upset if you read all this jargon only to find out the obvious, i.e. seasonal keyword is seasonal.)

What should be noted with great interest is that the second-most seasonal keyword in our set is “stuffed animals cats.” And unlike most of the other keywords in the set (whose highest search volume occur during the holiday shopping season in December), “stuffed animals cats” peaks in July.

Who woulda thought? :)

And while a global search volume of 5400 searches doesn’t make “stuffed animals cats” a short head term, it’s far from the extreme long tail, too.

Armed with this information, you can now proceed to time your client’s website optimizations and revisits so that peak rankings for “stuffed animals cats” will be reached around July, while the competition may not bother with looking into that keyword until the holiday season.

Like I said before: seasonal keyword research is all about getting your client’s website in the right place at the right time.

Hope you find this information useful, and I’d love to hear feedback in the comments on how this has (or hasn’t) helped your internet marketing campaigns.