Even Creatives Are Using Real-Time Data

I had the chance to speak with PointRoll CEO Mario Diez the other day about a McDonald's case study which involved turning big data into real-time intelligence. PointRoll worked with McDonald's agency H&L Partners to deploy "dynamic" creative messaging in real-time.

By using real-time data, creatives have a much better chance of delivering the right message to the right people at the right time. You already knew this, but it was funny hearing Diez explain it. "[Marketers] wake up every morning and they realize, 'Holy #%! my creative might be out of date. I just learned something this morning. I spent months building my TV assets and they might be out of date.'"

I asked Diez to define dynamic creative messaging. His response: "Creative messaging that is dynamically build dependent on the data of the marketer's campaign."



So, uh, that's still pretty confusing.

Here's the definition I gathered from the conversation: Changing the creative in ads based on real-time data.

The idea behind dynamic creative is that a single ad template can be placed and run like a traditional display campaign. The difference is that the content of that ad can change in real-time based on who is seeing it.

Using the "dynamic" real-time creative strategy, McDonald's saw overall campaign click-through rates (CTR) increase by 65%. The dynamic campaign had a CTR lift of 112%. Diez asked, "Why wouldn't you do that?"

Diez said that the actual placement of the ads doesn't change, just what happens after the ad has been placed. I also asked him whether or not the publishers care that the ads on their site can change (the advertiser is the same, but the content of the ad is what can change). He said, "There's an approval process for a lot of the publisher to be aware of what's happening on the creative side. When you have thousands of variations of ads, it just becomes unscalable." He added, "[But] we are at a point now where publishers have been working with media clients for a long time and there's a high level of trust."

To me, this seems like a good way to bring real-time data targeting to display advertising without dealing with the exchanges. More importantly, this seems like a good way to let the creatives get a feel for big data. Isn't everyone always saying that the best part of automation is that it frees up the humans to focus on strategy and creative?

2 comments about "Even Creatives Are Using Real-Time Data".
Check to receive email when comments are posted.
  1. Myles Younger from Canned Banners, August 2, 2013 at 4:17 p.m.

    I can sympathize with Mr. Diaz' somewhat obtuse definition of dynamic ad creative. It's not easy to encapsulate. Once ad creative is made to be dynamic, responsive, and contextually self-aware, there are a practically infinite number of tactics and applications. A tactic might be "let's allow the ad to look up the current weather in the viewer's area," whereas an application might be "now that our ad is enabled with weather data, let's have the ad feature products that are relevant to the current weather."

    I'll add one more thing to Mr. Diaz' quote "Creative messaging that is ... dependent on the data of the marketer's campaign." The creative messaging can also influenced by CONTEXTUAL information that does not come from the marketer (the current weather, for instance).

    At the end of the day, the pitch should not be that dynamic creative is some esoteric, must-have technology with a confusing 20-word description. Rather, the pitch should be that the dynamic creative vendor will assimilate the advertiser's campaign goals, and then work backwards to reduce an infinite array of options down to a simple mix of dynamic elements that will exceed the performance of static (non-dynamic) creative.

  2. David Haro from, August 5, 2013 at 8:58 a.m.

    What kind of real-time data are we talking about here? Dynamic creative messaging has been around for years. Based on where you've been and what you've done, a creative can be assembled accordingly. As Mr. Younger mentions in his comments, you can layer other data such as weather, even geo. The likes of PointRoll, MediaMind, Mediaplex, and Spongecell have been doing this for at least a good five years.

Next story loading loading..