Defining Your Marketspace Through Keywords

When I first started to apply competitive intelligence methodologies to search, I had to find a way to define and determine the left and right limits of not only the search marketspace but justify those limits with the clients' goals. With this in mind, I developed a theory. Not a theory as it refers to speculation or an opinion, but a theory in the scientific sense of the word to mean:

• Something that will explain and predict future observations of phenomena
• Something that describes a significant amount of observations
• Something that has few unknown variables
• Something that is provisional... the hypothesis may never be proven, but it can be falsified

My theory is that the keywords chosen as the core keywords for natural search efforts define your unique marketspace. Alteration of those keywords invariably change the marketspace and could redefine your true competitors (companies that have comparable or identical products or services), keyword competitors (competitors that don't necessarily monetize those keywords but could be blogs, wikis or other information/ resource related sites) and outliers (those sites that show up for the keyword results but seem to be un-related or random).

For the application of the theory, I will point to some research I recently did. I was researching the keyword marketspace for a company that provides STD testing. During our discussions, they wanted to show up for the following keywords:

HPV, hepatitis, AIDS, hiv pcr, hiv pcr test, herpes igg, herpes igm, std symptom, herpes symptom, hepatitis symptom, chlamydia symptom, gonorrhea symptom, hiv symptom, syphilis symptom, planned parenthood, aids hotline, std hotline, ASHA std, CDC hiv

While these keywords are good keywords for general STD information, they don't reflect what they do. These keywords dont define the company well, if well ranked, they wouldn't drive qualified traffic and they seem to reflect a desire for information, rather than testing. For these keywords, the top 5 competitors for this keyword selection (the top 5 are defined as sites that show up more often - discounting rank - in the top 15 in Google

Competitor URL’s	SE Presence	Search Saturation
www.cdc.gov	46	6.96%
www.ashastd.org	29	4.39%
www.nlm.nih.gov	18	2.72%
en.wikipedia.org	16	2.42%
www.medhelp.org	13	1.97%

Notice that each of these sites reflect what I predicted. Each site is for information, education and reference. None of these sites actually do any STD testing. The result of this information showed me that some more keyword specificity was needed. What if the keywords were changed to reflect more "testing" and "screening" related keywords? How would that change the search marketspace? So, I took the following keywords and looked at the marketspace:

chlamydia test, herpes test, gonorrhea test, hepatitis test, syphilis test, HIV test, AIDs test, chlamydia testing, herpes testing, gonorrhea testing, hepatitis testing, syphilis testing, HIV testing, AIDs testing, chlamydia screening, herpes screening, gonorrhea screening, hepatitis screening, syphilis screening, HIV screening, AIDs screening, confidential std testing, std screening, std test, std testing

Obviously, these keywords are much more applicable to the site's purpose and the overall goals of the client. However, what I found was slightly different. The sites that had the highest presence weren't sites that offered testing.

Competitor URL’s	SE Presence	Search Saturation
www.cdc.gov	37	4.07%
www.labtestsonline.org	36	3.96%
www.webmd.com	26	2.86%
www.ahrq.gov	24	2.64%
www.fda.gov	24	2.64%

Again, more informational and resource oriented sites... even LabTestsOnline.org is a non-commercial, peer-reviewed site about lab tests. I hit a wall at this point. It looked like I had falsified my theory. While the search space changed because the keywords changed, it didn't change to the benefit of the user or the comapny who provided the STD testing.

It's at this point where I started looking at the sites that didn't dominate the search results. I started thinking that the CDC was such a powerful resource, it essentially overwhelmed the search intentions of the user and forced its way into the results. So, what I did was compared the sites from both keyword lists.

I found that in the first keyword list, where the focus was not on testing or screening, cumulatvely, the companies that offered STD testing accounted for less than 2% of the total search marketspace. Looking at the second, more "testing" oriented keywords, sites that offered STD testing occupied just over 9% of the keyword marketspace. This is significant enough not to happen by chance.

The result was that after looking at the entire marketspace, companies that offer STD testing, in general, dont use SEO as one of their strategies. The result of the research showed that there was a distinct vulnerability in this particular market that had yet to be exploited by a smart company. However, to get over the .gov and .org hurdle, they had to be explicit and specific in their keyword choices. Otherwise, they'd get bowled over by the CDC like almost every other STD testing company that did show up.

When we use the checklist: Do these results provide any predictive power? Yes, they describe and identify both a vulnerability and a hurdle. Do these results take into account a significant observable size of data? Yes, the first one takes into account over 600 data points and the second on takes over 900 data points into consideration. Are the unknown variables known? Not really, unless I had access to Google's algorithm and a few PhD's... then I might be able to identify the variables, but the good thing is that these variables are all constant. The variables only change when the algorithms of the search engines change. Lastly, was the theory falsified? No, the data just wasn't where we expected it to be.

I would enjoy any and all feedback, this is something that I've thought about for a long time... and subjecting it to my peers for scrutiny will only make it better.