A/B Ad Testing on NE 39th & Sandy Boulevardby Anvil on July 17, 2009Pay-Per-Click (PPC)
On my way to work today, while stopped at a red light, my windows down because it was already 70 degrees out, the woman asking for money on the corner asked me if I could spare some change. I politely declined, but not before I noticed that in addition to the “Homeless. Anything Helps. God Bless.” sign she was holding in her hands, she had a second sign on the ground behind her. The second sign said “Can’t work due to injury. Even pennies help. God Bless.”
It struck me: this woman was A/B testing her cardboard signs! While I can’t be sure that this woman isn’t an underground PPC guru of some sort, I can be sure of this: we all should be – minimally – A/B testing our PPC ad text.
A/B is the simplest form of PPC ad text testing. There are all sorts of ways that you can get complex & technical with ad testing – multivariate testing using expensive and advanced algorithm based software, for example. But I’ve seen a lot of PPC accounts, and am surprised every time that I see an ad group with just a single ad creative that has been the only ad creative ever run in that ad group. I know, shocking, right?
Why should you A/B test? It’s easy.
- Increase your click through rate (CTR). If you are A/B testing two ads, and one is getting a better CTR than the other, you might want to toss out the one that is under-performing (see #2 below before actually dropping the hammer). Then, by adding a new ad to test against the existing high CTR ad, you are working to continually improve the CTR of your ad group.
- Leverage a strong conversion rate (CVR). You should, in fact, consider both CTR and CVR in conjunction when judging your A/B test. You may have an ad with a higher CTR, but that ad may convert worse if the ad text is misleading or poorly targeted. I could obtain an insane CTR with ad text that claimed something untrue or was too generally targeted (“70% off Shoes at site XYZ.com!” – when, in reality, site XYZ only sells one very specific type of shoe, for example.) The enticing ad text would get the clicks, but the visitor wouldn’t convert when they find that my website is a poor match to their query.
- Improve your Quality Score. In AdWords, advertisers with high Quality Score can be rewarded with cheaper cost per click (CPC) and improved ad rank. One of the main factors that influences Quality Score is CTR. By continually A/B testing ad text and fine tuning your PPC keywords and campaigns, you can help to improve your Quality Score.
A/B ad text testing is something that you should consider a never-ending part of your PPC advertising efforts. By testing at least two ad variations in conjuction, you will be able to determine which message is resonating with your potential customers and performing best for your business. Because, remember, even pennies help.