How Agencies Scale Winning Campaigns Fast

How to tell a real winner from a campaign that only looks good, and how to scale it: profit data in the feed, a tighter catalogue, and market-by-market plans.

Gamal

A winning campaign is one that grows the business at a profit you can live with. Plenty of campaigns look like winners on the platform dashboard and quietly lose money. Before you scale anything, make sure you are scaling the right thing.

Why Catalogue Campaigns Look Like Winners

For e-commerce brands with many products, sizes and colours, the best-performing campaigns on every platform are usually the ones built on the product feed: Performance Max and Shopping on Google, catalogue campaigns on Meta. They take most of the budget, and they report one combined number. A ROAS of 3, 4 or 5 for the whole campaign.

That total hides a lot. The algorithm promotes whatever sells, based on what each shopper seems interested in. It does not know which products make money. Meanwhile the commercial team has a clear plan: which products, colours and categories should carry the season, and why. The two rarely match, and nobody sees the gap while the headline ROAS looks good.

A Real Example: Under Armour

Before founding Gambra, our founder Mohamed Gamal led growth marketing for the regional group that runs Under Armour in the Gulf. The online store was doing about USD 15 million a year, and sales were rising while margin fell.

The campaigns were optimising for sales and ROAS. The team wanted them to optimise for profit as well. So they went back to the feed and added margin and profit data for every product. That gave the campaigns a second layer: as many sales as possible, at an acceptable level of profitability.

Then they started excluding products. Even high sellers came out of the catalogue campaigns if they were losing money on every order. On Meta, they stopped advertising the whole catalogue and showed the products and collections that fit the commercial plan. Advertising everything usually means spreading the budget too thin.

The signal came at about three months: profitability was up sharply on the same budget, with no extra spend. Over the period, new customers grew 2.4 times, acquisition cost fell 31% and profitability improved by 30%.

How to Spot a Real Winner

  • Judge it against your own sales and margin data, never the platform's reported ROAS alone.

  • Look at what it sells. A campaign carried by loss-making products is a cost, however good the total looks.

  • Give it time to prove itself. One great week can be noise. A pattern across several weeks is a signal.

  • Check it holds up when budget stays flat. If results improve without extra spend, you have found something worth scaling.

Scaling the Winner

Reviewing which catalogue products are driving profitable growth

Once a winner is proven, scaling comes in three directions.

More budget, raised in steps and tied to the target, so the algorithm and the creative keep up. How we pace this is covered in How Quickly Can Agencies Scale Your Ads?.

More creative. A winning idea needs variations, new angles and fresh formats before fatigue sets in. We look at what is winning and why, then build the next round of concepts on top of it.

More markets. Under Armour ran the UAE and Saudi Arabia in parallel, using the same profit-based method. The execution in Saudi was different in almost every detail: products chosen by local buying behaviour, different campaign types, an Arabic feed, different margins and creative that looked local while staying on brand. What works in the UAE does not automatically work in Saudi, and treating the second market as a copy of the first is one of the fastest ways to waste a winner.

What to Ask Your Agency

Ask how they decide a campaign is a winner. If the answer is the platform's ROAS, ask what they know about the margin on the products it sells. A good agency will already have that data in the feed, or a plan to get it there.