Kaizen Ad Performance
This is the second in a series of specific things you can do to improve the performance of your PPC account. As mentioned in the previous post, if you wish to improve the performance of your account, improve your ad text! Studying your ad space is a great way to see how your ads stack up against the competition, but you should also study your own ad statistics to see which ads are performing best, and which ones require improvement. This is loosely referred to as ad split testing, and there are a number of ways to perform it. The most important thing is simply that you do it continuously.
Choosing a Time Period for Ad Analysis
Generally when we perform ad split testing, we run an all time ad performance report. A longer time period results in more data. However, if we have refactored an account several months ago or revised ads based on a prior analysis, for example, we might find cause in some cases to choose a shorter period. Unless you have a good reason for choosing a shorter time period, choose all time.
Ad Split Testing
There are a number of ways to perform ad split testing, and this is explained separately. The bottom line is, don’t be too mechanical in your analysis. Your analysis should take into consideration the account structure (because slight variants of ads may exist across many campaigns and ad groups). It should also consider things like advertised specials, and other widely used ad attributes. If you can statistically prove that some ad attribute contributes to higher returns, then that is indeed very valuable information, not just with respect to writing better ads, but also for other marketing purposes. Any such findings should be particularly highlighted in our monthly Client Reports.
Don’t Revise Ads
Once you’ve done the analysis, and made a plan to revise ads, don’t!! Generally, we don’t revise ads, we pause them and create new ones. This is because Google doesn’t actually accept ad revisions; a revision is processed by Google as a deletion of the prior ad and the creation of a new ad. This is evidently so as to avoid trashing quality score calculations. Unless your account is bumping up against the limits on number of ads, In order to avoid losing valuable data just pause the prior ad and write a new one.
Conclusion – Ad Analysis
Document your work and findings in the Ad MD&A. Recommend improvements as applicable.