The scenario
The brand sells one category almost exclusively: air fryers, plus baskets, liners and a small line of seasonings sold as accessories. It launched two years ago as a single-product store and grew fast on TikTok and Instagram demo content — the crisp, the reveal, the same clip format that carried most of the category before it. The founder ran the ad account alone, learning Meta Ads Manager by trial and adding a new ad set every time an old one slowed down.
By the time White Friday approached, the account had built up more than a dozen ad sets across two campaigns, most of them variations on the same audience with slightly different age brackets, all bidding against each other in the same auction without anyone having noticed.
- Monthly revenue band
- pending client sign-off
- Average order value
- pending client sign-off
- Fulfilment
- Third-party warehouse, local courier, next-day across Kuwait
- Team
- Founder running media buying alone, one part-time content editor
Doubling the budget on White Friday, and watching every day get worse
One month before White Friday, the account had a genuinely good week: cost per purchase dropped, and the founder read that as room to grow. The budget doubled overnight across the existing ad sets, with no change to structure, creative supply or measurement. For the first two days, spend rose and results held. From day three onward, cost per purchase climbed daily, and by the second week of the sale it had drifted well past the point the founder considered sustainable.
The instinct was to blame the platform or the season. Neither was the actual cause. The account was running more than a dozen overlapping ad sets that had never been consolidated, most bidding against near-identical audiences in the same auction, and the same six creatives had been carrying the entire budget increase without a single new one added. Raising spend on an account already fragmented into that many pieces does not scale it — it just makes the same small audience see the same six ads more often, which is exactly what happened.
The account did not have a White Friday problem. It had a structure problem that a good week's numbers had been hiding, and doubling the budget was the exact move that exposed it.
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 fourteen ad sets into three
WinnerOne prospecting campaign, one retargeting campaign, one for returning customers, per platform
- Format:
- Account restructure, broad-audience prospecting with server-side conversion signals
The near-identical ad sets had been splitting the same small budget across more auctions than the account's daily order volume could ever teach the algorithm from. Consolidating let the budget accumulate enough signal in one place to actually learn, and frequency on the surviving prospecting campaign dropped within the first week.
Staging the budget back up in fixed steps
WinnerTwenty-five percent per step, held five days, rolled back after two days above the cost ceiling
- Format:
- Scheduled budget increases with a pre-agreed ceiling and rollback rule
The doubling that caused the original problem was undone the same way it happened: all at once. Stepping it back up in fixed increments, with a rule written down before the season started, let the founder tell the difference between a normal daily wobble and the account genuinely reaching its ceiling.
Adding Google Shopping as the second platform
WinnerCaught the price-comparing searcher who already knows the product name
- Format:
- Shopping feed, added only after Meta was consolidated and stable
Air fryers are a category shoppers search by name once they have decided to buy, which made Google Shopping a better second platform than a third TikTok campaign competing for the same discovery attention Meta already owned. It arrived with its own, separately tracked, marketing efficiency rather than being judged against Meta's numbers.
Pushing spend further once the ceiling was already found
LostOne more budget step, on the theory that White Friday demand would absorb it
- Format:
- A final step increase attempted mid-sale, against the account's own rollback rule
It lost, and it is worth publishing because it tested the ceiling directly. Return fell at the same rate it had before consolidation, regardless of season or creative, confirming that this account's growth had to come from average order value or a second market rather than more prospecting spend on a single-SKU category.
What we did — the optimizations, in order
Put server-side measurement in place first
We set up server-side conversion tracking on Meta and Google before touching a single budget number, with one naming convention for links so spend could be traced back to the platform that actually earned it.
Why: The founder had been reading platform-reported return, which had been inflated by duplicated conversions across the overlapping ad sets. Fixing measurement first meant every decision after it was based on a real number rather than a flattering one.
Consolidate the account before touching budget
Fourteen overlapping ad sets across two campaigns were merged into one prospecting, one retargeting and one returning-customer campaign per platform.
Why: Adding budget to a fragmented structure amplifies the fragmentation. Consolidating first is what let the next budget increase actually reach more people instead of showing the same people the same ads more often.
Confirm at least six proven creatives before any increase
We audited creative supply and found only two ads carrying meaningful spend. Four more were promoted from a small testing budget before the next budget step was allowed to go ahead.
Why: Two winners carrying a doubled budget is how frequency climbs and cost per purchase drifts up regardless of structure. Six creatives sharing the load is what let the account absorb more spend without exhausting the same small audience.
Budget in fixed steps with a written rollback rule
Twenty-five percent per step, held five days, rolled back one step after two consecutive days above the agreed cost ceiling, written down before White Friday started rather than decided in the moment.
Why: A rule agreed in advance is what stopped the founder from either panicking on a normal bad day or riding a genuine ceiling past the point of profit, because the decision had already been made before the emotion of a live sale arrived.
Add the second platform only once the first is stable
Google Shopping went live only after Meta's structure, creative supply and measurement were all confirmed stable, chosen because air fryers are searched by name once demand exists rather than discovered cold.
Why: Adding a second platform before the first is stable splits attention and budget across two unfinished jobs instead of one finished one. Sequencing it after stability is what let Google Shopping add net new demand rather than cannibalise Meta's.
Pre-load the season instead of reacting to it
White Friday and the pre-Ramadan air fryer spike were mapped four weeks ahead with a pre-agreed ceiling for cost per thousand impressions, so a normal seasonal rise did not get mistaken for the account breaking.
Why: Discovering a higher cost per thousand mid-campaign reads as a crisis; accepting it in advance with a ceiling reads as a plan. The difference is entirely in the timing of the decision, not the number itself.
Name the ceiling and change what growth means next
Once return fell at every budget step regardless of creative or platform, we documented the ceiling as confirmed and reframed the next quarter's growth plan around average order value and a bundled accessories offer rather than further prospecting spend.
Why: A single-category brand with a real ceiling that keeps feeding frequency is just buying the same customer twice. Naming the ceiling out loud is what redirected the founder's energy toward a lever that still had room in it.
What changed
The table above carries the numbers once the client signs them off, and the shape worth flagging now is that cost per purchase did not fall because the budget grew more slowly — it fell because the same higher budget was finally reaching a wider, better-measured audience instead of the same overlapping pool of people fourteen ad sets had been quietly competing over.
The Google Shopping addition behaved exactly as expected: it brought in demand that had already decided to buy and was searching by name, at a marketing efficiency tracked separately from Meta's, which made the blended number more honest rather than simply larger.
The genuinely useful finding was the ceiling itself. Confirming that return fell at every step regardless of creative or platform told the founder, in writing, that the next unit of growth in this account would not come from spend at all — it would have to come from the accessories bundle, which reframed the whole next quarter's plan before a single extra dinar was spent chasing frequency.
What we would do next
Build the bundled accessories offer — baskets, liners and seasonings sold as a set with the fryer — since the account's next unit of growth has to come from average order value rather than more prospecting spend on a single SKU.
Second, pre-load next year's White Friday and pre-Ramadan budget steps a full quarter ahead, with the ceiling already documented, so a good week never again gets read as room to double overnight.