How many is enough? How do you know which ones will work? Why do I get such different search counts? These are just three of the dozens of questions I get about keyword research. So, at the prompting of one of my newsletter subscribers, I decided to create a mini-tutorial of sorts for you to follow when doing keyword research, analysis and selection.
The first step in the overall process of finding keywords and phrases to use on your websites (or the sites of your clients) is to conduct the initial research. Basically this is the act of creating a list of possibilities. We’ll get to analysis and selection a bit later.
Regardless of which tool you use you’ll want to educate yourself about where that tool’s data comes from and whether it is clean. For instance, Google’s search tool is for its AdWords paid search program. And, as might be obvious, it comes strictly from Google and no other search engines. (They also have a search tool for YouTube searches.)
Their data is not clean… that is to say, automated searches are included in the results. Not good, in my opinion. Why would I care about the keywords automated software programs are searching for? I only care about humans.
Wordtracker and Keyword Discovery both offer clean data. They filter out the automated searches so your results only contain queries typed in by real people. Also, Wordtracker and Keyword Discovery both pull data from a variety of sources, not just Google.
When conducting my initial research, I start with a list of seed phrases. These are ones I get from a variety of sources including:
- The clients’ analytics program
- Keywords I would personally type in to find a site like theirs
- Google Insights
- The client
Make sure to have your settings geared toward people in the area you want to serve/sell. You also want to click “phrase match.” Do not conduct keyword research using the default settings of the tool (which are usually “global” and “broad match”). That will bring you search and competition counts that are far too bloated. (Check out this horror story of keyword research gone wrong.)
As I enter these terms into the keyword research tool, I’m looking at the suggestions it throws back at me. At this point I’m not too picky about what I’m adding to the list. Weeding through everything will come during the analyzing phase. If the suggestions are somewhere in the ballpark, they get put on my list.
Once I’ve gone down every rabbit trail I think might lead me to useful keywords, it’s time to turn a keen eye and narrow things down.
Typically what I’m looking for when I analyze keyword research is balance. Competition, search count and relevance/purpose have to combine in such a way that you end up with terms that bring in qualified traffic in enough quantities to make a positive difference on your site. It is unlikely (depending on several factors) that you’ll rank both highly and quickly for terms with a lot of competition.
In my opinion, relevance is almost always the priority.
In Wordtracker, you have the benefit of looking at their new and improved KEI formula. This, along with two other competition metrics, makes it a lot easier to determine the viability of each search term. Google and Keyword Discovery also offer their own versions of competition metrics, but they aren’t as detailed as Wordtracker.
After I’ve weeded through the initial list and removed any obviously irrelevant terms, I ask the client to go behind me and rank the relevance of each keyword from one to 10 (with 10 being the most relevant).
With this information in hand, I head to the keyword selection process.
Once the list is sorted by relevance, I look at the search counts and competition levels to determine which terms will work best on which pages. This, my friends, is a judgment call. Keyword research is enormously personal to each project. There is no way for me to give you guidelines that consist of suggestions like, “If the term has less than 100 searches per month, ignore it” or “If the competition level is above 65 look for another term.”
All of this depends on the page of your site the term will be on, your business model, where in the buying process the terms will be used and more. But you can use this as a guideline:
- Aim for the highest search count possible while taking relevance and competition into consideration. (EXCEPTION: niche marketing)
- Check for relevancy to: your website, your business model, the type of page, the buying process
A keyphrase may be relevant to your website as a whole, but not make a good choice for a Services page or a Category page because of the type of term. Or the keyword may be at the wrong point in the buying process and attract visitors who aren’t qualified yet.
If you have an advertising business model you might be looking for different types of keyphrases than if you have an ecommerce site. One is geared toward providing information and drawing people in who, perhaps, are looking for comparisons and reviews while and the other is geared strictly toward sales.
Balance, balance, balance. Aiming strictly for the highest search-count terms will lead you astray almost every time. When you give due consideration to all the pieces of the puzzle, you stand a better chance of creating a final list of phrases for each web page that work to bring in qualified traffic at the right point in the buying process. And that gets you one step closer to making the sale.
If you found this post valuable, others might, too. Please share (below) on your favorite social media sites. Thanks!