When AI Curates the Buy: Butler/Till’s Shortcut to Smarter PMPs

Butler/Till’s latest campaign for a financial services client didn’t just rely on traditional DSP smarts—it leaned into artificial intelligence to outmaneuver legacy ad tech inefficiencies. The agency tapped SWYM.ai’s SCaLE tool (Smart Curation and Learning Engine) to tailor Private Marketplaces (PMPs) that were more in sync with actual performance goals. The result? A 56% lift in conversion rate and a 26% drop in cost per conversion, all while reducing ad domain sprawl by 52%. That cut down on wasted impressions, particularly from underperforming MFA (Made-for-Arbitrage) sites.

What made the tool particularly useful was its ability to blend buy- and sell-side signals—from Google DV360 and Index Exchange, respectively—to continuously repackage high-performing impressions. It tracked which ad placements drove the most site visits and leads, then served lookalike inventory via daily refreshed PMPs. This hyper-targeting worked even within tightly geofenced ZIP codes, giving Butler/Till more of the impressions they wanted, not just what the DSPs were algorithmically inclined to serve.

For Butler/Till, this wasn’t just another AI toy—it was about reclaiming media strategy from a race-to-the-bottom bidding system. The approach avoided cheap reach from dubious sources in favor of quality impressions likely to convert. As Ryan Lammela, the agency’s group director of channel activation, put it, “We can take a little bit of control back from DSPs.” In an era where programmatic often feels like letting go of the wheel, Butler/Till’s AI-guided PMP strategy is more like installing a smarter GPS.

Full story at AdExchanger.