The Keyword Sweet Spot

How Keyword Research Shapes Your Optimization Plan

In the good ol' days of about 10 or so years ago it was generally enough to stuff the keywords meta tag, add some black text that included your keywords on a black background, publish the page and monitor results. We didn't worry so much about what people actually searched for as long as we could show the page coming up when we searched on those keywords in Altavista. Things have really changed and now it is extremely important to be in tune with selecting good keywords, or more accurately key phrases.

A number of years ago it became apparent that what was more important than just picking some words was to try to understand what people were looking for. There would be a "keyword research" meeting where the client would come up with a list of words they would use to search for their product. Even today, many SEO firms employ some variance of a keyword research meeting. While this may be good for getting those top search rankings for the terms the client likes, it doesn't mean there will be any increase in qualified traffic or ultimately conversions. It may sound cold, but in many cases I'm not really interested in what keywords the client likes, I'm interested in the terms the client's customers are likely to use to find their offering.

People Don't Buy From Themselves, So Internal Keyword Selections Are Often Irrelevant

Here is a real-world example. We did some work for a HR management software company many years ago when we still did keyword research meetings. The client had selected each of their product names and a trademarked term that they felt would become prominent in their industry in the near future. We dutifully optimized the site based on the terms they selected and were happy to see the site coming up #1 in the search engines for each of the terms within just a few weeks. Everyone was happy--for the first couple months. One day I got the dreaded "so where is all this traffic I was promised" call. The problem was very few people searched for the product names or the trademarked term. Plus, since all of these terms were very limited in use there wasn't much competition to beat for the rankings. We could have gotten to #1 without an optimization plan.

Think About How People Search

How people use a search engine is very important to keyword research. This is also one area where the search engines really make us in the SEO business earn our paychecks since the engines are constantly innovating to make the user experience more relevant. Typically people start with a common, though somewhat generic, term and then narrow down from there as necessary. For example, someone in Chicago who is looking for a divorce attorney may use terms similar to the following:

1. divorce
2. divorce attorneys
3. divorce attorneys chicago north side

The sooner the person finds what they are looking for the sooner they will abandon further search. As the saying goes, you always find what you are looking for in the last place you look. That's because once you've found it you stop looking. So what does this mean for keyword research?

The term "divorce" is likely too broad and too competitive for an optimization strategy. An easy way to tell if you are too broad is to search the term in Google and see if Wikipedia is the #1 result (which it is when I checked this one). The good news is if you are the divorce attorney on Chicago's north side none of your competition is likely to be coming up organically either. Plus, the searcher hasn't signaled their intent surrounding divorce at this point. Are they thinking about divorce, doing a research paper, divorcing a spouse or some other usage of the "divorce" term?

The searcher is now going to see that they didn't get any relevant results because they didn't signal intent. This is where they will start to narrow the focus using a secondary term such as "attorneys." Google, as well as most search engines, realize this happens so they leave the original searched term in the search box. This makes it easy for the searcher to add "attorneys" to make their next search: "divorce attorneys"

Now it starts to get fun. Since the searcher has now signaled intent, Google can start to personalize the search results. The goal of the search engine is to serve relevant content to the searcher in as few searches as possible. It is not likely that someone in Orlando searching for "divorce attorneys" would be interested in divorce attorneys located in Chicago, so the search results start to differ based on geography. This is why it is important to consider geography (or other personalization qualifiers) when conducting keyword research. Due to the size of the Chicago area there is still a strong likelihood that the searcher has not found a divorce attorney near them and will further refine by using an area qualifier. Contrast that with Orlando where the searcher is not likely to go any further since the metro area is smaller in Orlando than in Chicago.

Another wrinkle to consider for some sites is the number of mobile users anticipated. Searching via a mobile device is often done using voice recognition which naturally results in longer search queries. Think about a searcher looking for airport parking. If they are planning their trip from their computer they are likely to use something short like "parking near O'Hare" or "airport parking ORD." If they are in the car already the search term would likely be more natural language such as "cheapest airport parking near O'Hare airport."

Identify Keyword Sweet Spots

Unless you truly have something extremely niche, it makes sense to use one or more keyword research tools that give an indication of how popular a particular search term is. Google provides a free keyword research tool for their AdWords product that can give an indication of popularity. It is important to take the actual numbers with a healthy dose of salt, but the relative difference between terms should be accurate enough. This will tell you the patterns people actually use for search rather than what someone says they would search for.

It is usually best to start with a fairly generic term to see what similar terms people search for. Make a list of each of the relevant, similar terms, then start to deep dive on each one to find the sweet spot. The sweet spot will be different for every keyword and is basically the point at which each of the following conditions are met:

1. The keyword is qualified and relevant to the offering
2. There is significant search volume to drive the desired traffic level
3. There is little to no established organic competition for the term -OR- the term is a good value for PPC

Most sites can only be organically optimized well for 5-10 keywords depending on the competition and size of the site. This doesn't mean the site will only show up for 5-10 search terms. If a site ranks well for the sweet spot, it should also rank well for derivative terms. This is where it is important to use tools like Google Webmaster tools or an analytics package to understand what keywords searchers are using to find the site. Armed with this information the site can be continually optimized to take advantage of seasonal patterns, differences in language, etc.

If it is unlikely to rank competitively with just organic optimization, but the term is a good value for PPC, consider including it in a paid campaign. Just keep in mind that paid traffic must be highly targeted to make the investment worthwhile.

Sometimes there is no sweet spot for a keyword. This is okay. Continue monitoring the keyword to see if something changes to develop a sweet spot, but concentrate efforts on other keywords that will yield quicker results. As the overall site popularity improves it will become possible to effectively compete for keywords that once were out of reach.