I stumbled across Jason Cohen’s blog a few weeks ago and have been reading a bit of it lately. He’s got a great article today about Statistical Significance in AdWords campaigns (or any PPC split test for that matter). A great quote from the article…
Because the AdWords “click-through rate” is dependant both on the number of clicks and the number of impressions, you must not use “click-through rate” to determine statistical significance. Only the raw number of clicks can be used in the formula.
Read the rest at A Smart Bear.
A Smart Bear company logo
Agreed, CTR isnt an accurate indicator. Conversion ratio could be. May be including conversion ration into the formula, can make it more useful. Another metric could be what are the total number clicks that are generated by all advertisers for a given number of impressions, and % of clicks that we got at our website.