A/B Ad Testing Definition: A/B ad testing, also known as split testing or bucket testing is the process of testing specific elements within an ad. For example, you can test copy text, layout and images, etc. After you’ve researched your results from the ads you’ve run, you can determine what the most successful ad was.
A/B ad testing sounds guileless enough, right? However, when you are running, let’s say two ads within an ad group or within a specific Facebook campaign you should always allow them to alternate as consistently as possible between the two ads. Though Facebook has not completely perfected their advertisement platform, you should still be testing ads to determine which ads are best. You can test Facebook ads through the use of different images, headlines and text. After time has passed you should always be sure to evaluate your results and determine the best advertisement based upon the click-through rate (CTR) and the conversion rate (CVR). Or you can also try the combined portion of the two: click-through rate times the conversion rate (CTRxCVR). Testing ads is crucial because Google and Facebook love fresh ads and new content.
Seeing as every ad is different based on the product or service being promoted, you should also be aware of what determines the success of your ad. CTR and CRV are direct indicators of how an ad is performing. However, you will also want to measure your cost per conversion (CPC) – the cost of your redemptions, email sign-ups, likes and downloads, etc.
A/B ad testing is a valuable investigation yielding proactive expansions to your PPC campaigns. With various other PPC campaign strategies, A/B ad testing can certainly be ruined. On occasion, it can also damage your campaign performance. This is due to the over optimization of your ads or in other words: testing too many things at once. Over optimizing can sometimes result in the subsequent deterioration in either your campaign performance or the ad group. Of course CTR, CRV and cost per conversion are direct indicators of how an ad is performing; conversely there may be other needles of value to you and your ad, unequivocally.
Think of your ads as a sales pitch. Your ad is what brings your customers into the digital door – so to speak. One of the main trials and tribulations in running PPC ads is that you can’t fit ALL of your amazing messages and benefits into one single ad. Break up your most important announcements into individual ads or group some of them together. Once this has been completed you can then begin A/B testing.
The next component in the A/B ad testing process encompasses testing the best existing ad against a new contestant ad. If you are just starting out, it’s best to start with two advertisements to examine. Once the ads have gone to scuffle with one another, do your research and find out which ad had the best CTR and CRV.
Below are some of my core tips for great ad performance and evaluation:
Stay Creative – Copy-writting is not an easy skill to attain, exercise your abilities, research your competitors’ ads
Test your Images – Beware of iStock Photos (Specific to Facebook)
Run all Versions of the Ad – Like is said previously, it is best to begin with two ads, however once your prepared to move forward and test more than two ads be sure to run them all so your evaluations are even
Run your Ads Long Enough to Reach Data Significance – Without statistical significance, you gain nothing. Make sure that you are running your advertisement long enough to reach the data you need to make proper assessments.
Evaluate. Clean-Up. Start Fresh – Once you’ve gone through and researched the measurements from your first batch of advertisements, start fresh. Rebuild your ads with the statistical data gained from your ad tests and soon you have some of the best PPC ads ever written, until you have to test and write them again…