Do you ever get involved in ‘teaching’ things?
I sometimes do. Archery. Running. AdWords.
I dislike the word ‘teaching’. Partly because it reminds me of school, but also because it implies that I am going to tell you what to do.
Telling people what to do works okay with archery, where there are a defined list of things never to do. Like for example when an over-eager beginner will draw the bow string back to their ear, with no arrow attached. Gives me heart palpitations just thinking about it.
AdWords however is less clear cut. The problem with ‘teaching’ AdWords is that I am expected to stand on a podium at the front and decree “use this setting!” “But never use this one!”
Everybody wants to know which AdWords settings you should use. Whether you should target mobile devices. What bid modifiers to set. What ad rotation settings to use. Whether you should target the search partner network.
It’s like there is supposed to be some sort of magical optimal AdWords configuration that when set leads to guaranteed AdWords profitability.
What I can teach is how to approach your decision making in the right way. How to go about selecting the right campaign settings for the right circumstances. Unfortunately that requires thinking on your part, which for most people is less appealing than a defined list of do’s and dont’s.
One of the campaign settings I’ve been playing around with for a while has been the ad rotation settings. When you set up a new campaign, the default setting is ‘optimize for clicks’.
‘Optimizing for clicks’ is the default setting. AdWords is a clicks generation machine, and more clicks means more click revenue for Google.
There are however circumstances where you do want to optimise for clicks. In the early stages of an account for example, you won’t have enough data yet to optimise for conversions. So optimising for clicks is an obvious interim measure.
We’ve run campaigns recently where we have created as many as 40 ads in a single ad group. We’ve then set the ad rotation settings to ‘optimize for clicks’, and left Google to figure out which ads get the highest click through rates. If you look through our AdWords Survival Guide documentation that is what the process currently tells you to do.
In principle I am saying to Google “here are 40 ads I think might work. You go and figure out which ones are best”.
Do you want to know what happens when you do this?
What happens is that most of the 40 ads get very few impressions and under a handful of clicks. One or two ads get the majority of clicks.
The problem with this is that Google seem to have jumped the gun. They seem to optimise for clicks before they really know which ads are really performing best.
If ad A has had 2 clicks and a 2% click through rate, and ad B has 4 clicks and a 3.5% click through rate, you have no idea yet whether ad B will have a higher clickthrough rate in the long run. Really you need over 100 clicks on an ad before you can be sure of its long term performance.
Google however don’t seem to be concerned about that. (They just want their click revenue, dammit!)
So, I no longer suggest using ‘optimize for clicks’ or ‘optimize for conversions’. At least not in the beginning. Instead, I suggest you use ‘rotate for 90 days then optimize’ or ‘optimize indefinitely’, and follow our ad testing process every 90 days to evaluate your results.
Are there any other AdWords settings you find confusing? Let me know here, and perhaps we’ll cover them in future issues of there is sufficient demand.