Keyword research is the single most important aspect of any search optimization campaign, without exception. The keyword research process brings to light information that is both beneficial and directly applicable to many aspects of nearly any business. The actual keyword selection process must be undertaken with great care as each keyword represents an important business goal.
The most common criterion employed for keyword selection include:
- Search volume – The number of searches performed for a keyword.
- Competition – The volume and strength of competitor websites.
Generally keyword research is an attempt to find a balance between these polar attributes and usually results in some kind of compromise in terms of keyword quality or relevance.
Often the greatest difficulty of the keyword research process is finding a compromise that you can live with. This I believe demonstrates that the common approach to keyword research leaves much to be desired. The remedy is to be found by using new and innovative research techniques and more importantly by approaching the process with a renewed perspective based upon a more complete understanding of the gravity inherent in keyword selection.
Firstly we must abandon the idea of “picking” keywords. There are a finite number of search phrases that accurately describe your business; this process is one of discovery. Therefore a more accurate perspective is keyword “identification”.
Additionally the notion of “picking” keywords describes (accurately) the common research model used in the keyword research process: That is; mining every search phrase possible that is remotely related to one’s industry and then picking the best one’s from the list. This method is highly imprecise, involves minimal business intelligence and enervates resources.
Going back to the idea that each keyword represents an important business goal, we can see that “picking keywords” is a flawed research model. Can one select important business goals from a list of all possible business goals? Not if success or longevity are among one’s business goals.
I would say that a research model that more accurately reflects our “keywords as business goals” perspective would look something like this:
Keyword research has two distinct simultaneously occurring projects each with a different focus. They are:
- Business Intelligence Research
- Technology and Application
Business intelligence research is concerned with identifying information to address the following questions:
- What phrases best describe our business, technologies, services, corporate culture and mission statement?
- How well do these phrases describe our competition?
- What search phrases are used by our sales motivated target audience?
- What distinguishes this business from competitors? What are the phrases that reflect a consumer interest in those distinctions?
- Do our strongest real-world competitors also represent our top Internet competitors?
- Who would we like to consider our competitors? Is this the same as our actual competitors?
- What search phrases are driving conversions to top competitors?
Technology and application research is concerned with addressing the following questions and completing the following processes:
- In regards to competitors top conversion phrases; can our business become more relevant than our competitors for those search phrases?
- Are there any important search volume trends that can be observed for important keywords?
- Identify and compile an industry vocabulary containing all words, phrases and concepts that are commonplace within the business’ unique area of focus.
- Identify the search volume for phrases within the industry vocabulary.
- Exploration and alteration of the industry vocabulary using linguistic / semantic analysis.
For example mapping the following for elements of the industry vocabulary:
- Synonymy/Antonymy – Example: Light is an antonym of Dark.
- Similarity – Example: Warm is similar to Hot.
- Membership – Example: Senator is a member of Senate.
- Metonymy (whole/part relations) – Example: Piano has a part keys.
- Substance – Example: Lightning has substance Electricity.
- Attribute – Example: Mile and Kilometer are attributes of Distance.
- Causation- Example: Jokes cause laughter.
- Lateral bonds – Example: Happiness is related laterally to smiling.
Identification of pairs and triplets of phrases that most commonly occur together on top ranked competitor websites.
The formulation of two lists based upon information gathered from research data:
- A list of essential criteria which each keyword must fulfill in order to be considered (document co-occurrence, lateral strength with a set of core themes, etc.)
- A list of essential keywords identified during the research process based upon all of the information discovered.
At the completion of both research projects a series of collaborative presentation and brainstorming sessions should be facilitated so that the most effective keywords can be identified and implemented based upon the work of both research projects. Final selection would be based upon a variety of criteria, and keywords could be used for a variety of goals all of which is based upon solid data arising from the use of a thorough and intelligent research model.
The research model described above is of course very general and would need to be customized to individual businesses and unique business needs, however it effectively hints at the potential of a keyword research project that is based upon the perspective that keywords are important business goals.