The scenario
The brand sells seasonal gift boxes built around dates, chocolate and a rotating sweets line, sold direct to consumer through its own Shopify store with KNET, Tabby and cash on delivery at checkout. Two years old, one founder handling media buying alongside everything else, and a habit — inherited from how the account was first set up — of launching a fresh campaign every time a new box design shipped.
By the time we looked at the account, that habit had produced dozens of near-identical campaigns stacked up across two years, most of them still technically live and drawing a trickle of budget each. Nobody had gone back to turn the old ones off, and nobody had ever asked whether forty small, thin campaigns were actually cheaper to run than three large, well-fed ones.
- Monthly revenue band
- pending client sign-off
- Monthly ad spend before this engagement
- pending client sign-off
- Fulfilment
- In-house kitchen and packing, local courier with named delivery slots
- Team
- Founder plus three packers, seasonal hires before National Day and Ramadan
Forty campaigns, one budget, and a cost that climbed every time it was fed
The account structure told the whole story before we ever looked at a single result. Every box design the brand had launched since its first season still had its own campaign, most with two or three ad sets inside targeting nearly identical audiences, each holding a sliver of the daily budget. None of them had enough spend individually to leave the learning phase, and several were competing against each other in the same auction for the same customer.
The founder's instinct, reasonably, had been to spend more heading into National Day, the brand's biggest single week. Each time she raised the budget, she raised it inside this same fragmented structure, which meant the extra money mostly bought more auctions between her own campaigns rather than more reach. Cost per purchase climbed every single time spend went up, and the pattern repeated season after season without anyone connecting it to the structure underneath it.
This was never a creative problem or an audience problem. The account had forty accounts' worth of overhead and one campaign's worth of actual budget, and scaling it meant fixing the plumbing before touching the tap.
What we did — the creative that carried the spend
The losers are here on purpose. A test with only winners was never a test.
Consolidating forty campaigns into three occasion-based ones per platform
WinnerOne prospecting, one retargeting and one retention campaign per platform, replacing every product-level campaign
- Format:
- Account restructure, phased over two weeks to preserve pixel history
This is the change everything else depended on. Folding forty thin campaigns into three well-fed ones per platform meant each one finally had enough daily spend and enough purchase events to leave the learning phase and stay out of it, and it stopped the brand's own campaigns from bidding against each other for the same shopper.
Budget increases in fixed steps, pre-loaded ahead of National Day
WinnerTwenty-five percent steps every six days starting three weeks before the date, rather than a single jump the week of
- Format:
- Scheduled budget-step plan, agreed before the season started
The old pattern was a single large jump in the final week, exactly when every other gifting account in Kuwait was also raising spend and the auction was already at its most expensive. Stepping the budget up earlier, before the rest of the market piled in, let the algorithm adjust gradually instead of chasing a doubled budget on the busiest days of the year.
Confirming at least six proven creatives live before raising spend
WinnerA hold placed on the first budget step until the National Day angle sheet had six tested winners in market
- Format:
- Pre-scaling gate, tied to the creative-testing cadence rather than the calendar alone
In previous seasons the founder had scaled spend behind two favourite creatives because they were what she had ready, and frequency climbed within days. Holding the budget increase until six creatives were actually proven meant the extra spend had somewhere to go other than showing the same two ads to the same people more often.
Doubling the budget on the single best hamper creative during peak week
LostThe prior season's approach: put most of the National Day increase behind the one ad already performing best
- Format:
- Single ad set, aggressive budget concentration, the account's default move before this engagement
This is what the account had always done, and it is worth publishing because it is the instinct most small accounts follow. Concentrating spend on one winner drove frequency past the point a Kuwait-sized audience tolerates within days, cost per purchase rose through the week it mattered most, and the fix was never a better ad — it was more ads sharing the load.
A broad, untargeted retargeting campaign for National Day
NeutralRetargeting expanded to everyone who had ever visited the site in the past year, rather than recent, occasion-relevant visitors
- Format:
- Retargeting audience expansion test
Neutral rather than a clear loser: it did not damage cost per purchase the way the single-creative concentration did, but it also did not improve it, since most of the added audience had already forgotten why they visited a year earlier. The lesson was that consolidation and budget steps mattered far more than audience size at this account's scale.
What we did — the optimizations, in order
Put server-side measurement and one naming convention in place first
Verified server-side conversion tracking on Meta and Google, cleaned up link naming so every campaign, ad set and creative could be read consistently, and set blended marketing efficiency as the number the whole scaling plan would be judged on.
Why: Forty campaigns with inconsistent naming had made it impossible to tell which one was actually profitable before this. Clean measurement is what let us prove the fragmentation itself was the problem, rather than any single campaign's targeting.
Consolidate every product-level campaign into three per platform
Retired the product-level campaigns entirely and rebuilt the account as one prospecting, one retargeting and one retention campaign per platform, migrating the best-performing ad sets rather than starting cold.
Why: This is the single move that let the account's actual budget concentrate instead of splitting itself forty ways. Every campaign after this finally had enough volume to learn from and stop guessing.
Confirm six proven creatives before any budget increase
Set a hard gate: no budget step goes ahead until at least six creatives are performing under the target cost, feeding the newly consolidated campaigns rather than the old single-favourite habit.
Why: Extra spend behind too few creatives is what had driven frequency up in every previous season. The gate forced production to run ahead of the budget instead of behind it, which is the only order that keeps scaling from feeding on itself.
Raise budget in fixed steps, held and judged before the next one
Twenty-five percent increases every five to six days, held through that window and rolled back after two consecutive days above the cost ceiling, scheduled to start three weeks ahead of National Day rather than the week of.
Why: Starting the steps early meant the account was already stable and absorbing higher costs per thousand before the rest of the market piled into the same auction in the final week. A doubled budget judged after two days had been how every previous season's spike got discovered too late to fix.
Shape the National Day spend around the delivery calendar
Coordinated the budget schedule with the kitchen's production capacity and courier slots, so spend never outran the number of boxes the brand could actually pack and deliver on the days closest to the occasion.
Why: An account that scales orders faster than it can fulfil them turns a paid-media win into a delivery-slot failure, which for a gifting brand is the one thing that costs a customer for good. Budget and kitchen capacity had to move on the same calendar.
Watch for the ceiling rather than assuming it away
Tracked return on ad spend at every budget step, with an agreed rule that a second consecutive drop despite fresh creative and stable frequency means the account has hit its ceiling for the season, not a targeting problem to fix.
Why: Recognising the ceiling in the moment, rather than pushing another budget step past it out of habit, is what kept National Day week from repeating the exact pattern that had prompted this engagement in the first place.
What changed
The table above carries the figures once the client signs them off, and the shape worth flagging now is that spend rose into National Day week for the first time without cost per purchase rising with it in lockstep. That had never happened in this account's history, and it happened because the structure underneath the budget changed before the budget did.
The consolidation mattered more than any single creative or audience decision. Folding forty campaigns into three per platform is what let the account actually learn from its own volume instead of relearning the same lesson forty times in parallel, and it is why the pre-loaded budget steps behaved the way the plan expected them to.
The one losing test — concentrating spend on a single top creative — was worth running precisely because it was the account's old default, and confirming it still failed under the new structure closed the door on reverting to it out of habit next season.
What we would do next
Add Snapchat as a properly staffed second platform for the Eid and graduation windows, now that the Meta structure is stable enough to hand a second platform its own creative supply instead of splitting attention between both at once.
Second, build the National Day and Ramadan budget calendars a full quarter ahead each year, tied to the kitchen's production capacity from day one, so the budget-step plan and the delivery promise are set together rather than one chasing the other.