A/B Testing Display Ads - The Easy Way

March 17, 2016 Francis Dignan

How do you know that your ad campaign is the best that you can possibly publish? How can you be sure that the version you went for is the most engaging, creative, captivating display you are capable of? The only way you can be truly certain is by putting your ideas against each other through A/B testing, and thoroughly analysing the results.

Until relatively recently though, A/B testing in banner advertising was a painful, arduous task, and for many of you, it probably still is. Instead of the intuitive way you could A/B test emails or landing pages, for banner ads it would involve days, if not weeks, and entire teams of people.

Basically if you, as the marketer, wanted to change the copy in one of your ads, or you wanted to see if one image worked better than another, you would then have to go into your design program and redo your whole banner. This could involve coding, resizing text, and reanimating assets to ensure the ad maintains the same standard as the previous one. You’d then send this version of the ad to the publisher, who would then publish it for a fixed amount of time or maximum number of impressions, and you’d have to hope that it performs better than the first. All of this would cost a lot of time and money.

Fortunately, now there’s a better way to A/B test - the BannerFlow way. With this tool, you can easily make changes to existing banner ads, be it the size, visuals or content, without having to republish. Again, take the example of a retailer changing their campaign to see if it has any effect on engagement. All you have to do is enter the banner editor, change the copy as necessary, which is literally a case of deleting and replacing, and then you save it, without having to contact your ad network or republish, and then track the results.

Once you start using the scheduling tool though, the magic really happens, as you can run the two different campaigns concurrently, and analyse the results in real time. It looks like this:

Here, you can see that the yellow campaign is performing far better over a long period of time so that's the one to push.

It’s a nice, clean, simple way to see which of your campaigns are performing best, and thanks to the scheduling tool you can run them at the same time. You can even give each campaign a different weighting according to either results, or simply choose how often you want each version of the ad to be seen.

Also, if you decide you need to modify one, or both, of the live campaigns, you just pop into the editor, make the changes in a minute or two, save, and moments later your modified banners will beam all over the world, without ever needing to republish. It’s simply an edit, which will update the banners automatically. Then comes your real time analysis once again, which lets you see how your changes are affecting the  performances of your campaigns. Again, through the scheduling tool, you can weight one in favour of the other. You can set up one offer to start a few days after another, which is also a walk in the park. You just drag the desired campaigns to the desired dates, and sit back to watch the impressions roll in.

You can see here how clean and user friendly the scheduling tool is:

Here are 3 campaigns running side by side. From here you can change dates with a drag and drop, or pause a poor campaign.

Ease of use has to be a priority for A/B testing, as it’s such a fundamental function for marketers to use. Luckily, with this set of tools, it’s made super simple, which means you can look forward to analysing, modifying, and maximising engagement on every single one of your campaigns.

Are there any banner campaigns where you can think of a better B test to the A that you see? Tell us in the comments below!

Previous Article
Prepare your campaign for the Summer of Sports!
Prepare your campaign for the Summer of Sports!

Next Article
Banner Ads vs. Native Advertising
Banner Ads vs. Native Advertising