"Data-Driven Thinking" is written by members of the media community and contains fresh ideas on the digital revolution in media.
Today’s column is written by Michael Tuminello, senior director of product at Innovid.
It is becoming increasingly clear that data-driven creative is on the rise. For evidence, look no further than ad tech leaders like Annalect CEO Erin Matts, who recently came out in favor of “data and analytics … reuniting media and creative.”
A number of problems, however, need to be solved to make this happen. The first step from my perspective is to clear up the terminology used by the marketplace. Words like dynamic and targeting are used so broadly that they require immediate clarification.
While true that “programmatic advertising” is misleading in the sense that it generally refers only to media buying and not to the creative half of advertising, “programmatic creative” as a term has the disadvantage of being unwieldy, oxymoronic and abstruse all at the same time – quite an accomplishment for an industry that thrives on gobbledygook tech speak.
So what’s the difference? Targeting is the media decision – deciding to buy or not to buy a given impression based on data ingested from a variety of sources. Messaging, on the other hand, is the decision of what creative to show after the impression has been bought.
Demand-side platforms (DSPs) are the pulling the levers on the buying side, whereas messaging is handled by the agency ad server, and similarly operates off data ingested from a variety of sources or perhaps passed along by the DSP itself. The two components of targeting and messaging can work in tandem to achieve greater efficiency and results.
How so? A simple use case for how you might want targeting and messaging to work together is if you want to reach all 50 states with your advertising, but in each state you need to show slightly different messaging, such as different financing options for car payments, for example. Clearly you could do this entirely on the media-buying side, but you’d be dealing with 50 different tags in the DSP and you’d be buying geotargeted media at a higher price.
Instead you can send a single tag to the DSP, run on normal media, choose the right message on the creative side and still get back reporting for each individual message. Clearly there are also times when targeting does make sense – if, for example, there are separate budgets associated with each region – but the point here is that advertisers and agency can use both together and leverage each approach for its strengths as needed.
The general way to think about how targeting and messaging work together is that targeting draws the initial bucket of the audience you want to hit, and messaging then shows the appropriate creative for the individual viewer within that bucket, whether based on geo, previous behavior or ad views, audience subset or other data.
We will never achieve the goal of one-to-one marketing by decisioning purely on the buying side, and it’s also going to take a lot longer to get there if we keep having to stop to explain what we mean by “dynamic,” “targeting” and “programmatic creative.”
Let’s agree: In the context of programmatic advertising, targeting = buying and messaging = creative.