A few weeks ago, Hitwise released its latest findings on the number of keywords used in search queries across the engines. Not surprising, the number of searches involving three or more words is steadily increasing year over year, and the majority of searches include three or more words now. Why is that? I think it’s because as we all have evolved as search engine users, we’ve likely discovered that to get the most relevant result, we need to be as specific as possible in our search query. That being said, I’d expect this trend to continue going forward.
So what does this mean for your SEO strategy?
If searches on one and two keyword phrases continues to decline, that means less impressions and, likely, less clicks. Less clicks means less traffic. So if you’re focusing on 1-2 keyword phrases today, you’ll likely need to start retooling your strategy to include a few more keyword phrases with 3 or more keywords as those keywords will see an increase in searches, and likely traffic.
Before I started Search Mojo in 2005, I was working for an online survey company called WebSurveyor. WebSurveyor sold a basic online survey tool that could be used for multiple types of surveys. However, searchers were often looking for specific types of survey tools, such as “customer satisfaction survey software” or “employee satisfaction survey software” in addition to obvious keywords like “survey software”.
The challenge became that we had to begin providing pertinent content on the website around each of these survey types just to have a way to include the keywords in website copy and meet SEO content best practices. We found a legitimate way to do it, but it meant lots of new website pages and lots of content generation. And as we got further into the long tail terms, we found different ways to phrase terms like “customer satisfaction survey software”, such as:
Not to mention, we had to add the plural versions of those words, and all of these terms had a fair number of searches. In the end, even though we filtered and prioritized the phrases, our target keyword list for SEO was well over 100 words! It was crazy. But that’s the future of the long tail in SEO…. for every keyword you add to a phrase, you have to consider all of the variations of each term.
You may be saying to yourself, “OK, Janet. This all may be true for B2C or B2B product or service-related keywords, but is this REALLY going to affect EVERY type of keyword?”. And my answer is a resounding yes.
A few days ago, I was using Google Trends to analyze this very issue for a friend at National Geographic to see if searches on informational types of keywords also would drop. If you do a search on Google Trends for “African elephant”, you’ll see that searches on this term have steadily declined since 2004.
Does this mean that people just don’t care about the good old “African elephant” anymore? Not at all. Unfortunately, Google Trends can’t show long tail data because it does not have enough searches to do this analysis on most long tail terms. But I believe, and analytics data from several clients bears this out, that often, instead of a 1-2 keyword phrase driving the traffic, it is often associated terms that contain 3-4 keywords or more that are driving traffic.
In a recent analysis I did for one of our clients, we were trying to track back actual search engine traffic to the keywords we optimized their site for. Interestingly, more long tail keywords (associated with the 2-3 keyword phrase terms we optimized for) were driving site traffic than broader terms. It was not uncommon to find many various terms with only 1-5 visitors. But that long tail adds up, and search marketers need to consider that. Additionally, those long tail searchers had a lower bounce rate than those on 1-2 keyword phrase searches. And that’s what you want as a marketer — targeted traffic that sticks around (and hopefully converts!).
So hwo do you put it all together? Long tail sounds great from a traffic standpoint but a nearly impossible feat to achieve based on the sheer number of keywords you have to consider. Here’s how I suggest you tackle the issue, based on what I did at WebSurveyor with my insane keyword list: