The scenario
The brand sells bakhoor blends and oud gift sets to two different buyers on the same Shopify store: retail customers ordering one or two boxes for themselves or a wedding, and corporate clients ordering fifty to a few hundred boxes for staff and guests around National Day. The retail side runs through the normal checkout with KNET, Tabby and cash on delivery; the corporate side is usually arranged over WhatsApp and invoiced directly, which meant it had never once shown up in the ad account's own numbers.
Two years of growth had been driven almost entirely by boosting individual Snapchat posts rather than running structured campaigns, and every new bakhoor blend the brand launched got its own ad set rather than joining an existing one. By the time we were brought in, fourteen ad sets across two platforms were bidding on close to the same Kuwaiti audience, and nobody in the two-person team had a single number that combined online orders with the corporate revenue the ads were also quietly driving.
- Monthly revenue band
- pending client sign-off
- Average order value
- pending client sign-off
- Fulfilment
- Own stock, in-house assembly for corporate gift boxes, courier for retail
- Team
- Two founders splitting ads, fulfilment and corporate sales calls
A budget doubled on one good week, and a structure that never learned
The account we inherited told two stories depending on which number you trusted. The ad platforms reported a return that looked comfortable, and the founders' bank balance around the previous National Day told a different one: a week of strong-looking results had convinced them to double the daily budget overnight, and within four days the cost per purchase had climbed past anything they had budgeted for, while frequency on the core Kuwaiti audience nearly tripled because the same fourteen ad sets were still splitting an audience that had not grown to match the new spend.
The deeper issue was structural rather than seasonal. Every one of the brand's eight bakhoor blends had its own ad set, most running the same handful of boosted posts, all bidding in the same auction for the same eighteen-to-forty-five Kuwaiti audience. No single number told the founders what the account was actually doing, partly because Meta and Snapchat disagreed with each other and partly because the corporate orders — often the highest-margin sales in the business — were arranged on WhatsApp and invoiced by hand, invisible to every pixel in the account.
Scaling had amplified a structure that was never sound to begin with. The brand did not have a budget problem during National Day — it had fourteen campaigns competing with each other for a year, and the extra spend simply made the competition more expensive.
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 campaigns per platform
WinnerOne prospecting campaign, one retargeting campaign, one campaign for returning corporate contacts, per platform
- Format:
- Full account restructure on Snapchat and Instagram
Eight blends did not need eight ad sets bidding against each other for the same audience; they needed to be creative variants inside one campaign that could learn as a whole. Consolidating stopped the account from competing with itself in the same auction, and it was the single change that moved cost per purchase the most before any budget was touched.
Server-side tracking that includes corporate orders
WinnerA weekly reconciliation sheet that adds invoiced corporate orders to platform-reported revenue
- Format:
- Snapchat CAPI plus a manual weekly revenue reconciliation
Corporate orders were often the ones a National Day ad had actually influenced, and none of them were counted anywhere near the ad account. Once they were added to a shared weekly sheet, the blended efficiency number stopped understating what the ads were doing and gave the founders a real ceiling to budget against instead of a platform number neither of them fully trusted.
Doubling the budget overnight on one good week
LostThe exact move that triggered last National Day's cost spike, repeated once deliberately on a smaller scale to confirm the diagnosis
- Format:
- A controlled overnight budget doubling on the still-unconsolidated structure, run for four days before rolling back
Repeating the mistake on purpose, before the restructure, confirmed the diagnosis rather than assuming it: frequency climbed and cost per purchase followed within days, on the same fragmented structure that caused the original spike. It is the clearest evidence in the account that the problem was never the size of the budget increase — it was making that increase against a structure with nowhere sound for the extra spend to land.
Replacing boosted posts with structured conversion campaigns
WinnerThe same ad content, moved out of the boost button and into a proper campaign built on a conversion objective
- Format:
- Snapchat Ads Manager campaigns replacing post boosts
Boosting a post optimises for engagement, not for a purchase, and the brand had been paying for likes and shares while hoping they turned into orders. Rebuilding the same creative inside a conversion campaign let the platform's own delivery system find buyers instead of an audience that simply liked bakhoor photography, and it lowered cost per purchase on the exact same creative assets.
Adding TikTok as the second platform, budgeted by blended efficiency
NeutralTikTok introduced only after Snapchat was stable, with budget moved by blended efficiency rather than the platform's own reported return
- Format:
- New platform launch, discovery-focused bakhoor unboxing content
The founders had wanted a second platform for a year, mostly because a competitor was visibly active there. Waiting until Snapchat's structure and tracking were sound first meant the new platform's budget could be judged on whether it moved the same blended number the rest of the account was judged on, rather than becoming a second version of the same fragmented mistake.
Pre-loading the Eid and National Day budget four weeks ahead
WinnerA planned budget ceiling accepted in advance, with delivery shifted to evenings and a higher cost per thousand written into the plan rather than discovered mid-campaign
- Format:
- Seasonal budget and creative calendar built a month ahead of the occasion
The overnight doubling the previous year happened because nobody had planned for the season; the budget increase was a reaction to a good week rather than a plan. Building the ceiling and the step schedule a month ahead of Eid and National Day this time meant the increase in spend was already earning steadily by the time the occasion actually arrived, instead of scrambling to catch up with it.
What we did — the optimizations, in order
Put measurement on a footing that can see corporate orders
Server-side conversion tracking on Snapchat, one link-naming convention across both platforms, and a weekly reconciliation sheet that folds invoiced corporate orders into the same revenue figure as online sales.
Why: Without this the account was being judged on a number that missed its highest-margin sales entirely, and every budget decision that followed was being made against an incomplete picture. This is the step every later decision depends on.
Consolidate eight blends into three campaigns per platform
One prospecting campaign, one retargeting campaign and one campaign for returning corporate contacts per platform, with the eight bakhoor blends running as creative variants inside those campaigns rather than as separate ad sets.
Why: Fourteen ad sets bidding on the same audience were splitting the learning phase fourteen ways and inflating the cost of every one of them. One campaign per job lets the platform's delivery system find the best-performing blend on its own instead of the founders guessing which ad set to fund next.
Confirm creative supply before touching the budget
We checked that at least six proven bakhoor and gift-box creatives were live and stable before recommending any budget increase, converting the strongest boosted posts into proper conversion-campaign assets first.
Why: Adding spend to a fragmented set of boosted posts is how frequency climbs and cost per purchase drifts, which is exactly the pattern the account had shown the previous year. Confirming supply first meant the budget had somewhere useful to go.
Move the budget in steps, not doublings
Twenty to thirty percent increases held for five to seven days before the next step, with an automatic rollback rule after two consecutive days above the agreed cost ceiling.
Why: This replaced the exact move that had caused the previous National Day's spike. A step the account can absorb and hold is scaling; a doubling judged two days later on a structure that has not caught up is a bet, and the brand had already lost that bet once.
Add TikTok only once Snapchat is stable
TikTok was held back until the consolidated Snapchat structure had run cleanly for a full budget-step cycle, then introduced with its own creative and judged on blended marketing efficiency rather than its own reported return.
Why: Adding a second platform before the first is stable splits both the team's attention and the budget's learning, and it was the same instinct — chase the exciting new thing before fixing the foundation — that had produced the fragmented account in the first place.
Shape the season instead of reacting to it
Eid and National Day budgets, creative and delivery timing planned a month ahead, with an accepted cost-per-thousand ceiling written into the plan and corporate outreach scheduled early enough to close before the deadline crunch.
Why: The previous year's spike happened because the season was a surprise the budget reacted to rather than a plan the budget was built for. A brand whose demand is entirely occasion-driven cannot afford to discover its ceiling mid-campaign every single year.
What changed
The table above carries the numbers once the founders sign them off, and the shape worth flagging is that spend went up while cost per purchase held, which is the only combination that matters. It moved because the structure could finally absorb more spend without fourteen ad sets bidding against each other for the same audience, not because a single campaign found a lucky week.
Adding corporate orders to the weekly reconciliation changed how the founders read the account entirely. Several campaigns that looked mediocre on platform-reported return were quietly driving the corporate enquiries that closed at a much higher order value, and blended marketing efficiency is now the only number the team argues about, rather than two disagreeing platform dashboards.
The controlled repeat of the overnight-doubling mistake was worth running specifically because it lost. It proved, on this account and not in theory, that the previous year's spike was a structural failure rather than bad luck, which is what finally got the founders to commit to a step schedule instead of reaching for the same lever again this year.
What we would do next
Bring the corporate sales calendar fully inside the same planning cycle as the ad calendar, since the biggest orders in this business are still won on a phone call the ads merely start, and right now the two run on separate clocks.
Second, once TikTok has run a full seasonal cycle, revisit whether the brand has reached its ceiling on spend alone, since a gifting brand this occasion-driven may find its next real gain in average order value — larger corporate boxes — rather than in a fourth platform.