The challenge
Running performance marketing at scale across the UAE and the wider GCC region, Max Fashion had been leveraging TikTok Smart+ campaigns with TikTok's native All Products catalog feed — a setup generating solid results, but leaving one critical question unanswered. Was the default approach truly optimal, or could a curated, performance-led product set do more with the same budget?
The team needed rigorous proof, not assumptions. Specifically, they wanted to know whether focusing ad delivery on a smarter, more selective subset of the catalogue could meaningfully move the needle on key efficiency metrics — CPC, CPI, and CPA — without sacrificing the ROAS performance they had already established. The challenge was designing a test tight enough to answer that question cleanly, isolating the product set as the single variable while holding everything else equal.
THE SOLUTION
To settle the question definitively, ROI Hunter designed a controlled two-way split test that isolated the product set as the single variable, holding all other targeting and bidding parameters equal across variants. The test ran for six weeks with equal budget allocated to each experiment campaign.
The structure was kept as clean as possible:
-
BAU campaign running TikTok Smart+ with the native All Products feed served as the baseline benchmark.
-
Test Campaign with Two experiment ad sets ran in parallel with an equal budget split:
-
ROI Hunter test variant powered by a curated Bestsellers product set
- variant replicating the All Products approach as a controlled comparator
-
ROI Hunter's Bestsellers product set surfaces the most engaging and converting products from the client's catalogue based on live performance data, rather than delivering every SKU indiscriminately to TikTok's algorithm.
The underlying hypothesis was straightforward: presenting users with a curated selection of best selling, high-relevant products would sharpen audience engagement, lower media costs, and maintain or improve purchase efficiency without restricting scale so aggressively as to hurt volume.
Performance data was drawn from two sources throughout the test period:
- Branch for installs and purchase attribution
- TikTok dashboard for media metrics, giving the team a complete, cross-channel picture of how each arm performed across the funnel.
THE RESULTS
Six weeks in, the Bestsellers arm delivered a clear verdict. CTR rose 47% compared to the BAU All Products, confirming that the curated product selection resonated meaningfully more strongly with TikTok audiences. That engagement lift translated directly into media cost efficiency: CPC was 32% lower, a signal that TikTok's algorithm found the refined catalogue far cheaper to promote effectively.
The gains continued down the funnel. CPI fell 10%, meaning Max Fashion was acquiring app installs at a materially lower cost for the same spend. Most critically, these efficiency improvements also led to a 15% decrease in CPA compared to BAU, putting to rest any concern that narrowing the product set would come at the expense of revenue efficiency.
The test gave Max Fashion a data-backed foundation to restructure their TikTok Smart+ strategy around curated, performance-led product sets, and a repeatable testing framework for validating catalogue segmentation decisions in future campaigns.
