We all do it: take shortcuts. I mean, keyword research is such a pain in the ass that any way to short circuit the process is welcomed with open arms. The obvious short cut is to ask the client. He knows what he wants, he is an expert on market research, he has analysed all the sites that compete in his search space, he has sat down and analysed his AdWords campaign and culled the dodos, hasn’t he? 

What would you get if you were to just ask is pure gold, panned from his vast experience as a web marketer. So why the hell is he talking to you?

OK, end of sarcasm. Of course you should consult with the client about keywords, and the outcome will almost certainly pinpoint avenues for further research. However, the client is usually very close to the product or service he sells. He is an expert.  To expect him to understand or even relate to the generic terms often used during the early stages of the buying cycle might be a tall order.

In fact, no single source of keywords should be considered in isolation because each source will be slanted in some way depending on how the data is gathered. Nor should keyword research be considered a one off activity. Consumer trends change, technology-based products evolve at incredible speed, and people's searching habits are becoming more refined.

Keywords and Conversion

A lot of importance is placed on Conversion metrics, and rightly so. But at the end of the day, conversion and the bottom line are two entirely different things. For example, a person searching for a used car might use these terms:

• Used cars
• Used Ford
• Used Ford Mondeo
• Used Ford Mondeo Glasgow

Traditional SEO wisdom says generic terms ('used cars', for example) have more competition, are more difficult to rank for, and are the less likely to convert. Conversely, specific terms have less competition, are easier to rank for, and convert better.

The assumption is that because there are more competing sites, competition gets harder. This sounds true, but the devil is in the details. If I search Google for ‘used cars’, I can see there are 79,400,000 pages indexed that are deemed relevant for the query in some way.

(I am writing this article in Spain, so depending on which data centre you use you may see slightly different results.)

If you combine 4 of Google’s advanced operators (inanchor:, intitle:,  intext: and allinurl:) into one query, it is possible to pinpoint pages that have the term in their backlinks, in their title, on the page, and in their URL. The search then reduces the initial 79,400,000 to 11,200,000.

inanchor:"used cars" intitle:"used cars" intext:"used cars"

If you append an inurl:"used cars"

inanchor:"used cars" intitle:"used cars" intext:"used cars" inurl:"used cars"

The number drops to 3,760,000. This indicates that 7,440,000 of the pages could be index pages; however, this number also includes any URL that does not include ‘used cars’.  These numbers prove that some generic terms might not be as competitive as you first imagined. 

You can also look at country specific TLD’s:

inanchor:"used cars" intitle:"used cars" intext:"used cars" inurl:".au"

inanchor:"used cars" intitle:"used cars" intext:"used cars" inurl:".uk"

(To get this search to work you must include the speech marks.)

Searcher Intent

The keywords above represent a natural progression of the searcher as he refines his search and gets closer to his goal. There would be other steps in the process; those that moved the searcher from ‘used cars’ to ‘used Ford’, for example. Furthermore, there is no reason to believe that the entire process was completed in a single session.

The example terms move from generic to specific; they also give some indication as to how the searcher is progressing through his buying cycle, a process that could take days, weeks, or even months.

When you look at the terms, one thing that comes to light is that the intention to locate a used car is common to all of them, generic and specific. However, somewhere between ‘used cars’ and ‘used Ford’ something happened. Perhaps it was an online review, perhaps he discovered that Fords have exceptionally low maintenance costs and do 3 gazillion miles to the pint. So, if you sell and rank well for ‘used Saab’, you lost the sale before the searcher ever saw your wallet-liberating "buy one get one free" deal! Car manufactures are well known to most of us, so imagine someone researching the purchase of a product they new very little about, a speedboat for example. They will start by using very broad or generic terms, perhaps even ‘speedboat’. The first pages encountered, the ones that rank for generic terms, will guide the searcher towards more specific queries. Therefore, if you don’t target generic terms, the first pages the searcher will encounter will be those of a competitor,. Their brand will be seen and the information presented on these pages will influence every search from that point on.

Let’s Rethink This Conversion Thingy

A conversion could be a sale, a lead, an opt-in, an engagement (such as watching video, like Whiteboard Friday), request for brochure, or a contribution (this post on this blog, for example). There are countless others, but let’s consider an e-commerce site that measures site-wide conversion rate (i.e., purchase/visitor ratio).

What would happen if the webmaster of our e-commerce site deliberately targeted and gained rankings for some generic terms? Yep, conversion rates would plummet because more visitors would be referred from search engines. So, why would any webmaster on the right side of sanity deliberately target generic terms?

1. Influence the purchaser at the very beginning of the buying cycle
2. Build brand awareness at the earliest possible time
3. It is estimated that consumers visit a site at least six times before they make a purchase
4. Conversion as a percent of visits does not reflect the bottom line

Looking at conversions as a percentage of sales is a common practice, and I am not suggesting for one moment that you stop. However, e-commerce sites should also look at conversions as a monetary value (e.g., this month vs last month, this year vs last year, widget x sales this month vs widget x sales this month last year, etc).

The point is, targeting generic terms will probably lower purchase/visitor conversions, but that’s not necessarily a bad thing. It gives you the chance to educate, influence, and introduce your brand to buyers much earlier than companies you would traditionally consider competitors. Will this damage sales? On the contrary, remember intent; the searcher may not yet be in a position to buy, but you have introduced yourself at a very important point in the buying cycle...just make sure that first impression really counts.