Download the agency overview...
View our Agency Overview
Agency Overview

Expert Resources

More to Keywords Than Meets the Eye

If you’ve ever created a search analytics dashboard, you probably included a few rectangular slots dedicated to the “top 10 referring keywords”. It is such a commonly used metric that a dashboard seems incomplete without it. But, as an analyst, I question whether it generates any actionable insight, or helps to drive an effective SEO strategy.

Keywords are the life-blood of SEO, so it seems only necessary for them to be reflected in the analysis. However, analyzing data at the keyword level can be tedious, as the data sets tend to be massive and extracting this information requires post-processing, so much of it remains untapped. It’s unfortunate too, because this data can actually speak volumes about your visitors’ intentions, your search engine penetration, and your SEO campaign effectiveness.

Here are three ways to look at your keyword data and identify actionable insight.

Like Oil and Water

Your organic visitors consist of two groups of people: those searching for your products and services and those searching for your brand. These two groups behave differently, and thus should be looked at separately. Hence, the first step to better keyword analysis is to separate your keywords into branded and non-branded buckets (Using Excel). Just like oil and water, branded and non branded keywords do not mix well, and should be analyzed in silos.

Now that you have branded vs. non branded keywords, take a closer look at your overall keyword landscape. How much of it is branded versus non-branded? There are actually two values we can calculate to answer this question: the branded keyword share (number of branded keywords divided by total number of keywords) and the branded visit share (number of branded visits divided by total number of search visits).

In this example, you can see that while only 10% of traffic-driving keywords were branded, they accounted for nearly 30% of all search visits.

These metrics really shine when you add the dimension of time to the mix. For example, if you notice an increase in the share of non-branded visits over a select period of time, this could correspond to successful SEO efforts; on the other hand, an increase in non-branded keywords signifies more variety in the incoming searches, possibly reflecting wider search engine exposure.

Aggregate Over Similar Words

Keywords have a whimsical nature. A keyword is an instance of someone’s train of thought, and as we know, people tend to think differently (and sometimes, not at all). For example, while searching for suits, I might type in a search engine “buy suits online,” while someone else may try “online suit retailer”. The keywords differ, but the underlying search intent is the same – we both want to buy suits. Unfortunately, treating keywords like discrete and immutable entities is a perfect example of how we can lose the forest through the trees. There are large-scale patterns in your search data that you may never see if you differentiate between “suits” and “cheap suits”. A more informative approach is to aggregate all keyword phrases containing the word “suits” – regardless of other words that appear in the phrase.

The combination of exact and longtail keywords is going to smooth out the random fluctuations that individual keywords naturally exhibit.

Integrate Keyword Position Data

The longtail approach explained above is especially useful when combined with a second data source, such as your keyword position data. You can find your keyword positions by typing them into search engines manually, or by using one of a wide range of available tools. Since ranking higher in search engines for relevant keywords results in more traffic to your site, an increase in rankings should be reflected in the number of long-tail searches for that phrase. Several factors affect how sensitive your traffic numbers will be to a change in keyword rank, so an increase in search engine ranking may not tell the whole story. By correlating traffic numbers to the ranks, you can paint a more accurate picture of your keyword landscape: “’Cheap Suits’ rose from 13th to 2nd position in Google, increasing Cheap Suit-related traffic by 120%”.

Conclusion

Your organic keyword data can be a powerful resource. The key is looking past the individual data points and focusing on the big picture – the patterns and relationships in the data. Distilling this information may take extra effort, but the insights are well worth your investment.