DSP as coach, creative as the star: The new playbook for smarter ads

Back in the mid-2000s, I ran an old-school advertising network where we manually reviewed and approved every piece of creative, assessing how it would perform based on a gut feeling and some simple data.

With the rise of demand-side platforms (DSPs), however, we began to rarely think about ad creative. We focused almost solely on where the ad was running and who was seeing it. In other words, site and audience became everything, and the creative was simply an ad tag that we trafficked into the ad server.

Yes, we knew that Ad101 performed better than Ad102, and we could use dynamic creative optimization (DCO) to enhance performance. But we did not know why a specific ad creative worked, so we never got smarter over time on the creative side of the equation. In fact, the only time I talked about a creative was if it broke a publisher’s ad policy.

Now, technology has given us the ability to connect creative and media again. This synergy begins when your media is optimized to reach the right people at the right time in the right format — and your creative is perfectly crafted to be highly effective.

This is more than just DCO, which simply chooses between existing variations of an ad. Real creative data transforms how you make the creative in the first place and, even, what you make.This is where the magic happens. Think of it like a championship-winning sports team.

The DSP acts as the team’s real-time coach, deciding which player to put in the game at the perfect moment — and creative data is the general manager who builds the roster, ensuring every asset is designed to win before it ever hits the field. When you understand the underlying components of effectiveness — the pacing of transitions, the colors that hold attention, the moments that trigger action — you’re no longer just slotting players into the right moment of the game; you’re building a better roster that gives the coach even more options to choose from to win the game and to get better over time.

Reconnecting creative and media 

The key to unlocking this potential is getting creative and media teams on the same page. A recent study by the Winterberry Group found that 31% of companies say a lack of teamwork between these groups is the biggest challenge in proving their creative’s value. That is a problem worth solving.

So, what does this new approach look like in practice?

The answer lies in integrating DSPs and creative data platforms so that marketers can use creative performance signals to optimize programmatic campaigns. This isn’t just a distant idea — it’s the next logical step. Adding predicted creative performance to the bidding process at the DSP level adds a powerful new layer of context that goes beyond efficiency

It’s about delivering a more valuable experience for the consumer. It’s a win-win for all. 

Here’s how it would work in five steps: 

1. Establish creative scores: Before a new campaign begins, AI-powered technology analyzes a year or more of your historical data to uncover what drives results. It looks at creative and related performance data and calculates a score. Think of it as a creative “grade” on grabbing attention, provoking a response or driving purchases.

2. Integrate creative data into the bidding process: These creative scores can then be shared with DSPs to give the buying algorithm a head start. The DSP now begins bidding with a built-in understanding of each asset’s strengths, instead of waiting days — or weeks — for enough impression data to learn how each creative performs within a placement, audience or format. This shifts the DSP from learning by spending to making every bid more efficient from the first minute.

3. Optimize budgets pre- and mid- flight: With creative scores, the DSP automatically boosts the delivery of high-performing assets and reduces spend on weaker ones. The value isn’t just shutting off “bad” creative. It’s recognizing that performance varies by audience, placement and format. One asset may excel with upper-funnel audiences, another with retargeting, another on certain publishers or formats. The system routes budget to the right creative for each situation, not just the creative with the highest overall average.

4. Target based on contextual, audience and creative data: Contextual and audience signals tell you who is likely to engage. Creative data tells you what they are likely to engage with. Together, they produce smarter targeting decisions. If creative data, for example, shows that “close-up product demos” consistently perform better with value-oriented audiences, the DSP can automatically prioritize those assets when bidding into placements that are more likely to contain those users. This is targeting and creative selection working in concert, not in silos.

5. Measure creative impact to inform smarter decisions for future campaigns: With this model, post-campaign reporting becomes much more granular. When AI creates hundreds — or even thousands — of variations, you can test everything from background colors to call-to-action messaging. This is where true creative-media collaboration emerges. These insights flow back into the creative process, informing future briefs, templates and production decisions. Each campaign improves the next because the system learns not just what ran, but why it worked.

Creative data will fuel the next wave of generative AI

GenAI is fueling the next wave of efficiency, but it’s important to keep in mind that creative data is now fueling the next wave of GenAI.

Creative data is the training set behind AI-led creative design. Data is driving the reconvergence of creative and media.

Think of it this way: The coach (DSP) isn’t just calling plays in real time. They are also constantly receiving real-time performance data from the players on the court. Every score, every missed shot and every defensive play informs the strategy for the next game or the next campaign.

That’s the promise of integrating creative data directly into the bidding process: A true championship-caliber system, where the roster gets smarter, the coaching gets sharper and performance constantly improves.