Archive for September, 2006

Ad text quantity vs. quality

I received the following question the other day, which I would like to discuss in more depth:

What is the maximum amount of copies in and AdWords AdGroup?

The straight answer is “I don’t know.” I have never reached a copy limit within the same adgroup. But let me paraphrase the question slightly to highlight what I think the a good question to ponder about:

How many copies should I have in my AdGroup?

Purpose of ad copy optimization

When thinking about submitting new copies into an AdGroup and quantity is even a concern that comes up, it is worthwhile taking a step back and asking “What do I want to acheive by submitting many copies into the same AdGroup?”

The purpose of any copy submission or copy optmimization should be to find the one that performs best. But what does “best” really mean in the context of ad text optimization? Naturally, click through rate (CTR) is a good starting point. If you compare the ctrs of all your copies within the same adgroup over the same period of time , you might notice that some ad text has clearly higher ctr than the others. However, when you jufge strictly by ctr you might be making one of the common AdWords mistakes I have mentioned in a previous post.

Don’t forget the serving rate

Important information about the quality of your copy is the serving rate - if Google serves one copy more often relative to the others in that adgroup, it usually means that Google expects that copy to perform better. The serving rate, however, is not always dependent on the copy’s ctr or avg. cpc. I have seen copies with the highest serving rate in an adgroup that neither had the highest ctr nor avg. cpc. So good ad text is one that has a relatively high ctr and high serving rate.

Limit number of copies in one adgroup

Coming back to the quantity…it is not beneficial to submit many copies at the same time, because Google needs longer to determine for each one whether it is a good copy or not. In other words, Google’s ad optimizer kicks in later. This ad optimizer depends on an impression threshold for each copy, which I think is about 1,000. If your copies haven’t received 1,000 impresions yet, the serving rates are not being optimized and impressions are distributed equally across all your copies in an adgroup.

If you dump 20 copies into an adgroup, it will take you 20,000 impressions just to get the optimizer started. 20k impressions per adgroup can take quite some time to accummulate. However, if you only submit 3 copies at first, your adgroup only needs 3k impressions for you to know which one of the 3 is the best (this is not always true, of course, but bear with me here). Remove the 2 poor performing copies, and add 2 new ones. Benchmark the performance of the 2 new copies against the surviving copy from the previous test. Repeat. Repeat. This methodology ensures that you find a good copy quickly and you always benchmark new copies against it. You might be able to roll out the best copy to all your adgroups much faster - the lift to your campaigns can be quite significant.

Bookmark to:
Add 'Ad text quantity vs. quality' to Del.icio.us Add 'Ad text quantity vs. quality' to digg Add 'Ad text quantity vs. quality' to FURL Add 'Ad text quantity vs. quality' to blinklist Add 'Ad text quantity vs. quality' to My-Tuts Add 'Ad text quantity vs. quality' to reddit Add 'Ad text quantity vs. quality' to Feed Me Links! Add 'Ad text quantity vs. quality' to Technorati Add 'Ad text quantity vs. quality' to Socializer