Dynamic Creative In A Programmatic World

Dynamic Creative In A Programmatic World

As marketers get savvier at collecting and leveraging data, the opportunities around personalization messaging—also known as dynamic creative optimization—are evolving. In the last 10 years, the industry has reached the point where display advertisements can be dynamically assembled in a split second based on consumer data, context, and inventory availability, among other factors.

Most marketers agree that this highly tailored approach to advertising is here to stay. But now it’s imperative to understand the importance of the accuracy of the underlying data that drives the placement and the creative display.

Unifying Programmatic With the Creative

Dynamic creative allows marketers to deploy an infinite combination of creative elements in order to create highly customized advertisements. But if the targeting and the creative aren’t tightly aligned, an ad can easily miss the mark. It’s very important that the creative match the accelerated speed that ad buying and serving has achieved in the programmatic era.

Traditionally, creative teams haven’t been as involved in the technical execution of advertising. This has left a significant gap between those who conceive of marketing messages and design elements, and the technical teams or agencies that leverage data, execute ad campaigns and manage dynamic creative optimization.

But Ben Kartzman, co-founder and CEO of dynamic creative optimization platform Spongecell, sees programmatic and dynamic advertising capabilities moving closer to synchronization: “Over the next 12 to 18 months, I think executive creative directors will adopt that mantra and philosophy—and build briefs with programmatic creative in mind, rather than simply adding it on,” he says. If he is right (and I think he is) let’s examine the impact in some key areas.

Improving Retargeting with More Nuanced Segmentation

Retargeting is a commonly used practice that is often powered by dynamic creative: brands retarget people who visited their site, added something to their cart (but perhaps failed to check out), or demonstrated some other interest in the products or services offered. The retargeted ad includes a dynamically assembled presentation of the products (or products) the shopper just browsed, in the hopes of enticing them to buy or consider again.

The easiest retargeting route is to hit all browsers, product clicks and cart abandoners with a similar series of dynamic ads (featuring the products they recently viewed). However, some marketers can get much more granular about how they segment users for retargeting. For example, we found that by segmenting searchers (shoppers who searched their site for specific terms) and browsers (shoppers who browsed certain product categories), the search segment performed much better in retargeting campaigns.

We think searchers may be closer to actually making a decision and purchase, whereas browsers may just be exploring what’s out there, and not necessarily ready to pull the trigger. However, without tracking, segmenting and testing their retargeting campaigns based on shopper’s behavior and browsing patterns, companies wouldn’t have been able to optimize their dynamic creative campaigns based on these unique visitor segments.

Better Look-alike Modeling

The goal of retargeting is to get someone back to your site and convert them into a customer. But how do you reach someone who hasn’t actually abandoned a cart, clicked a product or even visited your site in the first place? Many marketers are now using dynamic creative to target new people who are in the right market for their product or solution, but may have not fully revealed it yet via their online activities.

This process, which most marketers refer to as “look-alike modeling,” involves using your own first-party data as a “seed set” audience, and then running a model across an outside, third-party set of data to identify “look-alikes” with similar attributes to your seed set. The new audience should include people who have never purchased from you or visited your site, but share similar same behavioral, demographic, attitudinal and/or geographic characteristics as people who have bought from you before.

Although many brands have already begun experimenting with look-alike modeling, the winners in the future will be those with richer data signals and advanced methods for segmenting their seed sets. These carefully segmented seed sets can be used to power more effective dynamic advertising campaigns, which are tailored based on the goals of the campaign and behavior of the audience.

Companies can even create different visitor segments based on “browsers” and “buyers” by separating out a seed set of high-propensity buyers, who are then modeled into a larger audience much more likely to purchase. On the other hand, brands that run a special seasonal deal could segment out a seed set of “deal-seekers” who would respond well to discounts and coupon codes.

Dynamic creative is only as powerful as the data behind it. The data not only tells marketers which ad units to buy based on who will see the messaging, it also helps determine what that messaging will look like. It’s an optimal situation whose goal is to serve consumers an ad that resonates and drives conversions.