KEI (keyword effectiveness index), which is defined in the next paragraph, is not as useful in SEO keyword research as it was years ago because it doesn’t factor PageRank and other non-textual strengths of competing web sites. This article explains this problem and what to do about it.
Readers familiar with KEI can skip this paragraph, which defines KEI. KEI is a number that represents the ratio of the quantity of searches done in the top search engines for a given phrase vs. the quantity of occurrences of that phrase in the search engines’ indexes. A higher number means that the probability of generating relevant search engine referrals to a page that uses the given phrase is higher because many people are searching for the phrase while relatively few pages on the web use the phrase. KEI is one of many metrics by which one can choose keywords to lace into web pages when doing SEO. When various keyword research tools, like WordTracker.com and KeywordDiscovery.com, use KEI, they refer to the count or frequency of the phrase within their databases that purport to reflect the quantity of searches done in the top search engines. The accuracy of these databases has long been a concern of SEO pros. This article explores the usefulness of KEI itself, even if KEI were 100% accurate.
In years past, before such off-page ranking variables as PageRank ascended, KEI was very useful in SEO. If matching searches to web page text is paramount, then the quantity of searches and competing web pages is key to maximizing search engine referrals. The more that web page rankings depend on off-page variables, like quantity and kind of incoming links or optimum CMS-SEO, the less important is KEI. However, KEI would seem to remain important in proportion to the percentage that on-page text is a ranking variable and in proportion to the labor investment required to generate positive returns. That is, if SEO copywriting determines only 25% of search engine positions but requires less time than other SEM and web marketing, then spending a little more time ascertaining KEI would seem worthwhile.
But a closer analysis reveals fundamental problems with KEI, no matter the valence of SEO copywriting within ranking algorithms. Before revealing this closer analysis, I’ll mention that users of WordTracker.com, including employees of my firm, have for years reported increasing instances of clearly inaccurate KEI (for example, KEI for many phrases being the exact square root of other data, as well as mathematically impossible negative KEI numbers). Also, SEO pros have justly doubted how accurately Google, Yahoo, and (now) Bing is represented by the databases of search frequency used by WordTracker.com and KeywordDiscovery.com, with many pros tending to prefer the more diverse, though un-revealed, portfolio of small search engines’ databases used by KeywordDiscovery.com. Businesses with larger budgets can get much more useful data than KEI alone by using tools offered by the likes of Hitwise.com and comScore.com, in addition to the tools offered by the search engines themselves. Describing the pros and cons of top research tools could fill a small book; here, I focus on KEI only.
The main problem with KEI is that it does not factor Google PageRank and other measures of authority and trust assigned to web pages. Today Google showed about 3,830,000 web pages containing the exact phrase, “financial advisors,” which would likely produce a low KEI. But what if all but a handful of the pages have very low PageRank? In that case, the phrase would likely be worth including in a keyword list even for a site with modest PageRank (the higher your web site’s PageRank, the more likely you will win using very competitive phrases). Of course in this case “financial advisors” is likely used by many web pages and web sites that have strong PageRank, so web sites with modest PageRank now and in the near future should use a long-tail phrase, like “independent financial advisors in western Massachusetts.” This longer phrase includes “financial advisors,” which most sites would not rank well for but still have a chance, while various longer and more winnable phrases are contained in that one long and most winnable phrase, e.g., “financial advisors in western Massachusetts” and “independent financial advisors.” The point is that KEI does not provide a measure of the actual competition for a phrase because the mere quantity of competing pages indicates nothing about the strength of the competition.
Of course good keyword research examines many factors other than KEI, regardless of the usefulness of KEI. Relevance to the site and page, conversion rates, average gross profit per converted search, projected future search trends and past seasonal trends, and other variables should be considered during keyword research. Keyword research is one of the most demanding tasks in SEO, requiring many months of training and experience. Again, adequate coverage of the topic of keyword research exceeds the scope of this, or for that matter any, one article.
So, should one consider KEI, and if so how and when? It depends on available budget, potential ROI of SEO, and how many pages on the web use the given phrase. Regarding the latter, selecting among variations of a very niche term appearing in few web pages, like “magnetic levitation conveyance,” could be helped by KEI because no matter what the PageRank of competing sites, a relatively high KEI variation would be worth selecting. Regarding potential ROI of SEO, if a web site could earn huge ROI by top placement, then it’s worth spending the time to use KEI in conjunction with other data in order to select the most productive phrases.
The time required to use KEI in conjunction with other data pushes up the cost of keyword research, which leads to the third criterion for deciding when to use KEI: available budget. Keyword research follows a variation of the 80-20 rule: the last 20 percent of perfection can take 80% more time. If your potential ROI and your order fulfillment capabilities are sufficiently high, then it’s worth finding the funds to factor KEI, or even to spend the tens of thousands of dollars on the more expensive keyword research tools mentioned above. If the funds aren’t available, and you can’t convince upper management or clients to pull funds from other marketing budgets, then you’re probably best off using KEI only as a tie-breaker among semi-finalists in your keyword list, or disregarding KEI altogether. In that case, Google’s and Bing’s keyword tools alone may suffice.
Chances are, if you don’t possess or can’t countenance the funds to consider KEI in connection with other keyword data, your equivalent competitors won’t either. By the same token, if you allocate less resources to keyword research than you should, your competitors may not make the same mistake, and you could lose in the search engine marketplace.