The challenge
Reserva already ran a sophisticated performance-marketing operation, with an established set of filters distinguishing Best Sellers from Poor Performers across its catalog. But those segments were manually curated and refreshed on a schedule, not in real time — which meant the team's ability to react to shifting product performance was only as fast as their next manual review. Reserva wanted to push the intelligence of their segmentation further: make it more dynamic and more responsive, with the goal of squeezing more Meta ROAS out of the same catalog and the same budget.
THE SOLUTION
Reserva's team had chosen ROI Hunter's AI-driven Smart Segments, but before shifting a core part of their strategy, they wanted proof. They asked for a controlled, side-by-side test: Smart Segments versus their existing business-as-usual segmentation, under equal conditions, to see which approach actually drove better results on their Meta campaigns.
The test was designed to isolate one variable only: segment quality. The team built two standalone campaigns of equal budget - Control group, running Reserva's existing hand-built segments, and Smart Segments, running ROI Hunter's AI-powered segments — with identical targeting, creative, and optimization settings. Both cells promoted the full catalog rather than a narrow slice of it, avoiding catalog fragmentation.
The core of the Smart Segments cell was its Poor Performers segment. Unlike Reserva's manual rules, which classify products on a fixed cadence, Smart Segments continuously reclassifies the catalog using live performance signals - spend, conversions, and revenue as they happen, so a product sliding into underperformers gets flagged and deprioritized immediately rather than weeks later.
As the test progressed, this turned out to be the single biggest driver of the result: the team observed that Smart Segments' handling of Poor Performers, more than any other segment, was what pushed the account's numbers up over the manual baseline. Segment Prioritisation was deliberately excluded from both cells to isolate segment quality alone
The test ran for three weeks, with both campaigns launched on the same day and monitored throughout to confirm spend was tracking evenly and neither cell was skewing on volume, placement, or audience.
THE RESULTS
By the end of the three-week window, Smart Segments had delivered a 17% Meta ROAS uplift from switching to AI-driven segmentation alone, with no other variables changed.
The uplift traces back to how aggressively Smart Segments cut waste: spend on Poor Performers dropped 75% as the AI continuously reclassified underperforming products and pulled budget away from them in real time. That reallocation showed up everywhere — CTR climbed 17% and CPC fell 8%, evidence that Meta's delivery system had more efficient inventory to work with once the weak performers stopped soaking up spend.
For Reserva, the result settled the question the team had set out to answer: AI-driven segmentation outperformed their established business-as-usual process, not just in theory but with their own catalog and their own budget on the line. Smart Segments is now becoming the account's default strategy, with the team planning to extend dynamic segmentation across the full account.
