The more time I spend in the SEO game, the more fascinated I become with semantics. Semantics is the study of how people and machines use and interpret words. A significant portion of my job (or any SEO’s) is looking at products on websites and then researching key words that customers are likely to use when searching for those products. That may seem pretty easy but it is truly amazing how many different words people use to convey the same meaning. It is even more amazing that search engines are getting so damn good at knowing what we really mean when we search for something. Knowing how both people and search engines refer to your products is the most important part of any SEM or SEO strategy.

If you sell blue widgets and want to show up on the term “blue widget” then you would include “blue widget” in the content of your site and maybe get links from other sites using that term as the anchor text. This is all well and good and your boss or client will be thrilled when you are #1 on Google for blue widget but where do you go from there? You can start by using stemming variants or variants that are based on the same root word. This can be as simple as using the plural form (blue widgets) or maybe even different prefixes or suffixes.

The next step is to determine as many synonyms as you can for your target term. Search engines provide suggestion tools to help with this (and there is always the thesaurus) but you also want to talk with key personnel in your organization who deal with customers. This is especially important if you are in a niche market where the search data maybe lacking. This is when you find out that customers in the Northeast call blue widgets blue bobbits.

Once you have a healthy list of variants and synonyms, you will need to prioritize according to popularity. There are many tools to help with this. Google Insights is one of my favorite free ones. If you participate in any paid search advertising then you likely have some good data of your own as well.

Google has been making some great strides in semantic search and specifically in parsing synonyms. Recently Google engineer Steven Baker wrote a great post about synonyms in search and for the first time actually gave some stats on how good Google is at extracting synonyms from their vast amounts of search data.

… our measurements show that synonyms affect 70 percent of user searches across the more than 100 languages Google supports. We took a set of these queries and analyzed how precise the synonyms were, and were happy with the results: For every 50 queries where synonyms significantly improved the search results, we had only one truly bad synonym.

Steven is not one of the more public Googlers (see Matt Cutts) but I definitely look forward to reading future posts from this guy.