Advertising: Dynamic Creative Optimization (DCO)
If you are reading this article, you are most likely aware that you can generate dynamic creatives with the Storyteq platform in no time. And if you are using dynamic creatives in your campaigns, you probably want to know how you could get the best results. Therefore, we have put some learnings, tips and tricks for dynamic creative optimization in this article.
Dynamic Creative Optimization
Let’s start with the definition of Dynamic Creative Optimization (DCO). DCO is a (display) ad technology which creates personalized ads based on data about the viewer at the moment of ad serving. Because the creative is both more relevant and also tested and optimized, dynamic ads typically outperform their static counterparts.
If you want to find out which creative works best for which audience, it is important not to test more than 2 variables at the same time. The more variables you put in 1 test, the less reliable the outcome of your test will be. A tip: make hypotheses. For example, creative A (video 1 with voice over X) will work better than creative B (video 1 with voice over Y). This way you’ll get trustworthy and clear learnings out of the A/B test you’re doing.
A global rule of thumb: start with A/B testing the most noticeable aspects within the ad. In this way, you can iterate to the best performing version step by step. In general, the more noticeable aspects are earlier in your creative (first 5 seconds), are in focus and in the centre of your creative.
Make sure the dynamic creatives that you’re testing are “live” for a period that’s long enough. To get results you can actually extract valid and significant conclusions from, we recommend you get at least 1000 impressions per ad-variant, preferably more to make the results even more significant.
Ideally, during the test period, there are no other campaigns or marketing activities going on that interfere with the test. Of course, we realize that for most company’s this isn’t a possibility, but it’s good to at least take this into account while setting up your test your dynamic creatives.
There are several outcomes you can base your test on. For example clicks, conversion or View Through Rate (VTR). When the VTR is an outcome you want to optimize on, we would advise testing this with ads that are longer than 10 seconds, especially on YouTube. Ads shorter than 10 seconds are often skippable and/or too short to measure a true VTR.
In short: make sure you test one variable per A/B test, start with the most noticeable variables in the video, make sure to get enough views per creative, try to downsize other marketing activities during the test period, and make sure the video is long enough if you want to measure the effect on the VTR.