The ROAS playbook I use for GCC e‑commerce in 2026
What’s actually working for D2C and retail brands across the Gulf right now — creative cadence, audience math, and the one budgeting trick I keep coming back to.
few weeks ago, a founder in Dubai messaged me at midnight: “Our ROAS just collapsed. Same creative, same audiences, half the return.” The honest answer: the playbook from 2024 is dead. Here’s what’s actually working for D2C and retail brands across the GCC right now.
The new ROAS reality
The number you saw on Tuesday is not the number that matters. Attribution windows, iOS noise, and the slow death of view-through have made dashboard ROAS roughly 40-60% of the truth.
Here’s what we do instead — we triangulate. Three numbers, every Monday:
- Platform ROAS — what Meta or Google reports.
- Blended ROAS — total revenue ÷ total ad spend.
- Marginal ROAS — what the last riyal of spend returned.
The dashboard isn’t lying to you. You’re just reading it like it’s 2021.
Creative cadence beats creative quality
The brand putting out 14 mediocre ads a week is beating the brand putting out 2 great ones. Not because mediocre is good — because the algorithm learns faster from variety than polish.
Our cadence at Addox: 3 hooks tested every Monday, 2 formats per hook, 1 long-form Friday, 0 reposts of last month’s winner.
The 60/30/10 budgeting rule
Split every account three ways: 60% proven core (no tinkering), 30% active testing (worse blended ROAS expected — that’s the cost of finding next core), 10% wild bets (the founder thinks they’re stupid; the wins fund themselves 20×).
Brands that plateau are usually 95% in core, 5% in testing, zero in wild.
Closing
The brands winning in the GCC right now aren’t the ones with the biggest budgets. They’re the ones running a system, every week, without flinching. None of this is glamorous. All of it works.