Canvas manages a large automotive OEM as a non-endemic advertiser on the Amazon DSP, with over 1,000+ campaigns running per year across two teams. At that scale, the optimizations that actually move the needle—granular bid adjustments, pacing checks, frequency tweaks—were constantly at risk of being deprioritized due to human bandwidth constraints.

Removing the Bandwidth Bottleneck

Canvas partnered with Gigi to build full-funnel agentic workflows that codified their best practices, scaling across their campaigns and teams. Workflows now cover Amazon DSP pacing checks, frequency adjustments, bid optimization, budget shifts, and inventory exclusions across STV, OLV, and display campaigns.

Gigi generates recommendations for the Canvas team to approve with a single click and can execute changes automatically while logging every action. Reporting that once took hours now takes minutes. With 100+ creatives live at any given time, Gigi also monitors for DSP rejections and resubmits them automatically so nothing falls through the cracks. Across both teams, 20+ workflows were deployed.

The Results

Within weeks of launching the bid adjustment workflows—custom adjustments across geo, device, and domain based on historical KPI performance—performance improved sharply:

• Off-Amazon CPA decreased by 52%

• Off-Amazon conversion rate increased by 116%

On efficiency, the two teams reclaimed 60 hours per week combined:

• Team A: 9 workflows → 20 hours saved per week

• Team B: 11 workflows → 40 hours saved per week

That time is being reinvested into testing frameworks, strategic client guidance, and thought leadership—the work that actually requires human expertise. The result isn't just a more efficient operation. It's a team that can optimize with a level of consistency and precision that simply wasn't possible before.