Data Choices
We compile. You decide.
Search Engine Optimization Experts!
See the video! Click the camera!
 |
 |
 |
 |
 |
 |
Keyword Effectiveness Index - KEI
The Keyword Effectiveness Index, or KEI, can be used to estimate the relative effectiveness of a group of keywords. In its most basic form KEI is simply a measure of a terms popularity, judging from a recent history of searches over a given period of time, and its competitiveness as determined by the number of responses returned by the search engine. There are certainly some shortcomings to the KEI, and they are most pronounced when we look at Pay Per Click, or PPC advertising. The KEI does not take into account the cost associated with bringing individuals through to the website. While this is always an important factor, it can be extremely important for new businesses. As companies begin to "test the waters" with a select group of keywords, they cannot simply presume that the phrase that brings the most visitors to the site is the best phrase to utilize with PPC. Unless, of course, the company has no concerns about the cost of the campaign, which is not likely. Data Choices introduces a concept that takes into account the Keyword Effectiveness Index as well as the cost of bringing paid hits through to the site: the DCI, or Data Choices Index.
Data Choices Index or DCI
The Data Choices Index, or DCI, is designed to bridge the gap between organic hits and simple ROI analysis. Simple ROI analysis is too simple, in our opinion, and a lot of money is wasted by not properly analyzing ROI for PPC. Nevertheless, that is another topic for another time. Most (if not all) companies simply compare the cost of bringing visitors through to their site via a specific keyword. Then they count the revenue created by that term and they have the most simplistic ROI available for PPC. If PPC costs exceed that of revenue generated by that term, the term is discontinued. Conversely if ROI exceeds the PPC cost, then the company keeps the term active and re-evaluates over a given period, usually monthly.
Let's ignore all the flaws in that strategy and go straight to the problem of the newly formed company that has not yet built a strong enough presence to judge ROI at any level. These companies need a method to determine which terms are working for best for them in an economical fashion. For instance, the term "real estate Internet marketing" may have been entered into a specific search engine 154 times over a recent period. But if that term has 89,700,000 results returned, then we can easily see it is a highly competitive term. KEI is calculated by squaring the number of searches and then dividing by the number of results and then multiplying by 1000. In the above example this would be 23,716 (154 x 154) divided by 89,700,000 and then multiplying that result by one thousand to get .26439. Obviously this result is time and engine specific, meaning that you need to utilize similar time frames and the same search engine in order to be able to compare results in a meaningful manner. It seems the click through rate to cost per click ratio would be an important factor in any PPC campaign. This can help you determine where to spend advertising dollars and can be an important part of your Internet marketing strategies.
DCI
Let's presume that in the above example your data shows that it costs $6.25 on average to bring people to your website using this keyword. And let us further presume that this represents a 1.6% click through rate. Ideally we would like the click through rate to be as high as possible (CTR should increase as the term moves higher on the list, but so will CPC, or cost per click). So what we really want is the CTR to be high and the CPC to be low, relatively speaking. Therefore DCI takes the CTR (1.6 - taken as 1.6 and dropping the percent) and divides it by CPC, or 6.25 in this case. This result (.256) is then multiplied by the KEI x 1000, so we get .256 x .26439 x 1000 = 67.68.
Now let's take a look at "real estate search engine optimization" as our second keyword. Let's presume our search shows 108 searches over the same time period and the same engine, and it returns 8,290,000 results. The KEI is (108*108)/8290000 = 1.407. This appears to be more than five times as effective as term one above! If our CPC is $15.62 and that CPC brings us a response of .75% (3/4th of a percent) we get .75/15.62 = .04802, which we multiply by KEI times 1000 or 1407 to obtain a final DCI of 67.56, or nearly the same!
If we merely looked at KEI we might be tempted to run only keyword two, given that it is five times more effective. And if we only looked at CTR/CPC we might be inclined to only advertise term one, given the click through rate to cost ratio. But by utilizing the Data Choices Index, which is a combination of the KEI, the click through rate and the cost per click, we can better see that these terms are amazingly similar in the cost to results ratio. In this example we would likely conclude to either run both terms, or neither term. If we have a table of thirty, forty or even fifty terms but also have a limited budget, we can use the Data Choices Index, or DCI to better determine where our advertising dollars are best spent!
Other Criteria:
Are there any minimums? Like minimum number of click throughs before using DCI? Yes. First there should be a minimum of 27 searches for a specific keyword in order to determine a valid KEI. There should also be a minimum of 27 click throughs before determining a valid CTR, or click through rate. If you do not have the required minimum you should lengthen the time frame to allow for these numbers. But remember to keep time frames consistent across terms that you will compare!
DCI calculator (default info is from keyword one above)
Please note that this calculator does NOT look up results by keyword. Use Yahoo Search Marketing or any other PPC campaign you are running to fill in the number of searches, results, CPC and CTR rates! If you just need KEI info only click here!
Get FREE updates to this article and related articles!
back to top see the article on Arrivenet!
