Ad Copy Split Testing Made Easy

Posted on October 3rd, 2008 in Promotion

If you're tired of refreshing this page every day like a d-bag, you should probably subscribe to the RSS feed.

Ok, hopefully everyone by now is tracking their ppc campaigns with prosper202, and if you’re not, you’re a retard and should start doing so immediately.  P202 makes it easy to see which of your ad copies is giving you a better overall ROI, which is very important. It’s also important to always be at least making progress with your campaigns, either through improvement of a current campaign, or by making fresh ones.  Improving your ad copy is something that should be done at least weekly, and possibly daily depending on your volume.  Nickycakes makes it a point to make fresh text ads at least once a week for all his campaigns.

So, what’s the best way to go about split testing ad copy and improving on it?  Well, it’s pretty simple.  When you make a campaign, always make 2 ads per ad group.  There are two schools of thought on this one.  Some people think you should make the ads similar with only small changes to give you a better picture of what specific changes do to affect ctr.  Nickycakes, on the other hand, makes every ad very different.  This gives you more opportunity to find great ads you may not have expected to do well.  Pretty much every successful campaign the Cakes has had has been him submittting some campaign on a whim not really expecting it to work and then coming home a few hours later to find the budget completely spent and a couple G’s profit in the ol’ affiliate account.

Ok, so you’ve made 2 ads for your adgroup…now what?  Well, now you wait for a while to see which one does better, then you dump the one that was worse and come up with a new one to put in its place.  There are 2 things you can look at to determine which ad is “better”, CTR and ROI.  You can use p202 to individually determine which ad has a better ROI, which is certainly interesting and useful info, but in terms of scalability and long term profit, CTR is what you should be paying most attention to.  This is, of course, assuming that you’re not stupid and all your ads are on topic.  Making an ad that says “YOU’RE A GIANT DOUCHEBAG” to advertise your free scholarship offer may get a decent CTR but will probably not leave you with many leads.  So yeah, you look at the CTR, because a better CTR, at least for search and contextual advertising, generally means your ad will be placed in a more visible spot for less money than your competition.  This means more traffic for less money which means you get to go buy yourself that hot new xbox 360 game at the end of the month.

So how do you know if you have gotten enough traffic to make a good decision on which CTR is actually better?  Cakes’ buddy Charles dropped a twitter bomb a couple months ago about using splittester.com for split testing.  Basically you take the lifetime # of clicks and lifetime CTR for each ad, enter it in the little boxes, and it’ll tell you how certain you can be of which ad is gonna be the better performer in the long term.  If it’s over 95%, it’s time to make a new ad and get rid of the worse one.

Making new ads is something that would be good to set aside a time every week to do, like say monday or friday every week you devote to making new ads.  It’s all about staying ahead of your lazy ass competition.  Just remember: Adapt or get paved the fck over.

Keep it real.