Advertising technology has completely changed the way that marketers can target and customize their ad messages. An influx of consumer data, from online and offline sources, has changed the way that marketers can find an audience online. But because this market segment has grown rapidly in recent years, marketers haven't focused very much on finding the right person to like. Matching the right person to the right creative message has become secondary.
Dynamic creative optimization and versioning have made it so that marketers can customize messages depending on a number of variables, including the consumer's data profile, but also the weather, time of day, location, and past consumer behaviors. But all those factors push advertisers too often to do too much. I've written in the past that many marketers can produce too many variations -- sometimes in the thousands -- completely eliminating the ability for them to learn anything from their dynamic creative efforts. This also restricts their ability to change their minds and add or remove messaging on the fly or introduce new messages when available.
And while too many creative versions are by far the most common misstep when it comes to delivering real-time creative, there are other common challenges as well. Fortunately, many of these have simple solutions. Let's address the problems and how marketers can work around them to get maximum benefit out of dynamic creative.
Trying to optimize against too many variables
As outlined above, there are so many variables available in dynamic creative optimization that it's really easy to get carried away. The rule of thumb for all marketers should be that that if you're measuring and optimizing everything, then you're measuring and optimizing nothing. Limit yourself to the factors that are most important to your KPIs (more on this momentarily) and identify the biggest drives of success.
No clear key performance indicator
Many advertisers use real-time marketing strategies to spit out hundreds of versions of an ad to see what works, and what it tells us. But without defining success upfront, there's nothing to learn. Brands need a clear goal to drive the optimization process forward, whether that's conversions, video views, or online purchases. Once those factors are defined, the team can address which creative permutations drive the highest success rates. The whole point of creative optimization is to serve creative that repeatedly hits goals. That's impossible without clear goals.
Not enough impressions
Optimization is a great tool, but it requires investment in media to truly take hold. Small sample sizes don't allow the technology to learn prohibiting proper optimization. Marketers should use the first 10 percent of the campaign (impressions served) for the first optimization effort. If it's a small campaign, then 10 percent might not deliver enough impressions to collect any meaningful insights. Major changes in performance across a small sample size are hard to replicate. In some cases, you could be looking at a statistical fluke, rather than learning about the audience or creative.
A/B testing and expecting optimization
Of course there's a difference between optimization and versioning. Optimization is about learning what works and letting technology push forward better performing permutations, while A/B testing and versioning focuses on delivering creative based on understood pre-defined parameters. Both tactics take advantage of real-time creative, but are successful at different accomplishments. Versioning is successful when an advertiser already knows about their audience. Optimization is best if you want technology that learns and changes the campaign as it progresses.
Dynamic creative is a great tool that lets marketers deliver customized messages to different audience segments in real-time. Finally, marketers can easily deliver ads on a one-to-one basis in display, rather than rely on the one-to-many that is persistent across so many other media outlets. If marketers are patient and take the time to avoid the pitfalls, they'll reap the success of delivering better creative to audiences that they understand on a much deeper level.
Brian Goss is chief technology officer at PointRoll.
On Twitter? Follow iMedia Connection at @iMediaTweet.