We cut spend by 16% and ROAS rose to 3.73, with a 1.97% CVR over 7 months

No playing with numbers: we spent 16% less and sold more, and ROAS rose to 3.73. It came from very precise work.

The problem

A store in Kuwait with good sales but unstable numbers. In June we raised the budget by 25%, sales grew only 8%, ROAS fell from 3.31 to 2.79, and cost per order climbed from $24 to $28.6. More money going out for a worse result.

The strategy

We stopped and decided we wouldn't change anything on gut feeling again. Decisions come from the numbers, not from mood.

The execution

Daily tracking of every dollar. A sheet closed every day (spend, orders, cost per order, ROAS, AOV). As soon as the data came together, something nobody had noticed showed up: there were fixed days every month with no sales, and we were spending on them as if they were peak days.

Reallocating the budget. Higher spend on the paydays when the data says people buy, and lower spend on the slow days. Exactly the same budget, but every dollar gets spent on a day that sells.

Improving creative instead of starting from zero, and locking in the campaigns that work (Evergreen), because any edit sends a campaign back into learning, and you're the one paying for that learning. Finally, CRO on the product page until the Add to Cart rate went up.

The result

July made $50,593 in 29 days (the highest month in the store's history) on $12,949 in spend, 16% less than June. ROAS went from 2.79 to 3.73, with 642 orders (+23%), and cost per order dropped from $28.6 to $20.9.

Over 7 months: $339,680 in sales, 4,389 orders, and a 1.97% CVR, roughly double the market average. If the same visits had converted at 1% like the market, we would have lost half of those sales on the same ad money.

Before you ask "should I raise the budget?", ask: what am I converting at?