Adding a 5th Dimension to PPC Testing
Over on one of our partner sites, Intelligent PPC, there’s a post today about an A/B testing process that really surprised me with how simple, yet profound it is. By doing a simple copy-paste in Adwords Editor, and playing with some ad scheduling, you can add day-parting to your landing page tests. It’s so easy I’m surprised everyone doesn’t do it.
PPC analytsts are generally familiar with a standard set of landing page test dimensions – placement of elements on the page, voicing the call to action (color and wording), varying degrees of branding, and ultimately, the conversion event itself. You can set up basic split-testing by simply duplicating entire campaigns, changing one test dimension and analyzing the results. Most people test a landing page’s physical elements – color, wording, content, form or no form, etc. Very few people test these different elements along with timing.
The example on Intelligent PPC focuses on a campaign that lands a user on a page with a video element – the question was whether the video should be click-to-play or auto-play. The answer was that it depends on the time of day of the visit. How do you address that question so that users during the day aren’t assaulted with an auto-playing video? Easy. Create two identical landing pages, one using auto-play, one not using auto-play. Use the ad scheduler so that the click-to-play video landing page campaign runs during the day, then when it goes off, turn on the other campaign that uses the auto-play video. You can also combine this with geo-targeting if you’re a national advertiser, so that west coast users won’t get the autoplayer at 2 in the afternoon and get themselves in trouble at work.
Time-slot testing gives you a new level of insight into behavior and campaign efficiency. It’s worthwhile to remember that aside from the traditional testing variations, you can test time with your landing pages using ad schedulers in combination with various landing pages.
Sorry, ouija has no special extra insight.
If you enjoyed this post, please consider leaving a comment or subscribe to the feed and get future articles delivered to your feed reader.
Interesting point, bet many of others like myself are having a “duh” moment for not thinking of conducting such tests. I am interested in seeing some results on this if you do conduct such a test. Would content, calls to action or any other factors change in the day-parting?