The scenario
The brand sells multi-surface cleaner, laundry concentrate and refill pouches in Kuwait, direct to consumer, on a Shopify store with KNET, Tabby and cash on delivery at checkout. One founder, a part-time warehouse hand, and a catalogue built almost entirely around bundles, because a single bottle sitting next to the co-op's shelf price never wins that comparison on its own.
Demand follows the calendar hard. Orders roughly double in the month before Ramadan, when every household does the deep clean before the first ghabga, and spike again after the first big dust storm of the summer, when the majlis floor and the car seats need exactly the demo the brand's own ads were supposed to sell. What it did not have was a way to make a new ad that was not another bottle photographed on a marble counter.
- Monthly revenue band
- pending client sign-off
- Average order value
- pending client sign-off
- Fulfilment
- Own stock, local courier, heavy-carton delivery across Kuwait
- Team
- One founder, one part-time warehouse hand
Four bottle shots, one promise, and a cost per purchase going the wrong way
When we tagged the previous quarter of spend, the pattern was obvious: every ad was a version of the same clean bottle on a clean counter, promising a stronger clean than whatever was already under the sink. Frequency on the main prospecting audience had climbed past the point a Kuwait-sized audience tolerates, and the brand's response had been to add budget, which only pushes the same ad in front of the same tired eyes faster.
The audience was not the problem. The pitch was. Nobody scrolling past a household-cleaner ad in Kuwait doubts that cleaner cleans; that was never the question stopping the sale. The real hesitation was arithmetic and trust: is this actually worth more than the identical-looking bottle at the co-op, and does the refill pouch really hold up the way the first bottle did.
That hesitation was sitting, unanswered, in two years of WhatsApp messages and product reviews the brand had never mined for advertising. The account did not have a targeting problem or a creative-quality problem. It had never once put the co-op comparison or the refill question in an ad, and it was spending more to be ignored by the same audience faster.
What we did — the creative angles
The losers are here on purpose. A test with only winners was never a test.
The real mess, not the clean counter
Winner"This is what my stovetop looked like after the dust storm."
- Format:
- Vertical, 15 seconds, phone camera, unedited before-and-after on a real stovetop
Every previous ad opened on a clean surface, which asks the viewer to imagine a mess she is not currently looking at. Opening on the actual grease and dust closes that gap instantly, and it reads as evidence rather than as an advertisement. It won clearly on hook rate and held its cost per purchase as spend increased, which none of the studio-shot creatives managed.
The co-op comparison, said out loud
Winner"Same bottle as the co-op. Here is what's actually different."
- Format:
- Vertical, 18 seconds, two bottles side by side with the price per litre written on screen
The most frequent unspoken objection in the brand's own reviews was that the co-op sells something that looks identical for less. Naming that comparison first, then breaking down the price per litre once the refill pouch is counted, turned the objection into the argument rather than something the brand hoped nobody would think about. It performed best on the retargeting audience, where the co-op comparison is exactly what stalls a saved item.
Safe around kids and the family cat
Winner"My toddler crawls on this floor twenty minutes after I mop it."
- Format:
- Vertical, 16 seconds, filmed in a real living room with a child in frame
The second most common WhatsApp question, and one the brand had never answered anywhere public: whether the residue is safe for children and pets on the floor shortly after cleaning. Answering it inside a real home, with a real child in the shot, converted better on the parents segment than any efficacy claim, because it removed a worry rather than adding a promise.
The influencer glamour shot
Lost"My cleaning routine as a busy mom."
- Format:
- Vertical, 22 seconds, styled influencer content shot in a staged, spotless kitchen
A reasonable hypothesis that lost, and worth publishing because it lost for a specific reason: the kitchen in the shot was visibly staged, and the category's whole credibility runs on the mess being real. It pulled decent views but almost no purchases, because the exact audience that responds to grease-and-dust honesty is the audience that disengages the moment a cleaning ad looks like a lifestyle ad.
The eco-friendly claim
Neutral"Clean your home without the chemicals."
- Format:
- Vertical, 14 seconds, plant-and-water B-roll with text overlay
Flat across every platform. It was not the wrong claim, just the wrong priority: nobody in the brand's own message history had ever asked about ingredients, only about strength, price and safety around children. Leading with a value the customer had not asked for outperformed nothing and pulled focus from the angles that actually addressed her real hesitation.
The refill maths, spelled out
Winner"The refill pouch is where the bundle actually pays for itself."
- Format:
- Vertical, 17 seconds, text-on-screen breaking the bundle price down per bottle refilled
This is the arithmetic that never appeared anywhere on the product page or in an ad before this round. Once a shopper can see that the third and fourth refill are what make the bundle cheaper per litre than a single co-op bottle, the decision stops being emotional and becomes a spreadsheet she has already done in her head. It became the strongest bottom-funnel creative in the set.
Speed of the demo, not the science of it
Winner"Sixty seconds. No scrubbing."
- Format:
- Vertical, 12 seconds, real-time countdown overlay on a stained car seat
A second objection-led test built to check whether the stovetop winner was a fluke or a pattern. It performed nearly as strongly, which confirmed the pattern: speed and effort, not chemistry, is what a Kuwaiti household actually buys a cleaning product for. It also worked well as a TikTok-native format because the countdown gives the platform's audience a reason to watch to the end.
What we did — the optimizations, in order
Tag the back catalogue before making anything new
We exported the previous quarter of ads and tagged every one by angle, hook type, format and language, then worked out what share of spend had sat on creatives below the target cost per purchase.
Why: Every ad in the catalogue turned out to be a variation of the same clean-counter idea, which meant the brand had never actually tested anything — it had produced the same creative six times with a different bottle angle. Naming that first stopped the next round from repeating it.
Build the angle sheet from the inbox, not a brainstorm
We read two years of WhatsApp messages, product reviews and return reasons and pulled out the seven questions that came up most, most of them about the co-op comparison, the refill and safety around children and pets.
Why: The brand had spent three years answering the co-op question in one-on-one WhatsApp replies and had never once put the answer in front of the cold audience that was asking the same thing silently before scrolling past.
A two-week production sprint in real Kuwaiti settings
Twelve pieces shot over two weeks: a real stovetop after a dust storm, a car seat, a living-room floor with a toddler in frame, filmed on a phone in real homes rather than a studio.
Why: A category whose whole credibility rests on a real mess cannot be shot in a set that looks staged. The studio-shot influencer piece proved the point later by underperforming, but the production plan was already built around avoiding it.
One testing campaign per platform, frozen everywhere else
One campaign per platform, one ad set per angle, equal budgets, five to seven days, with the bundle offer, the audience and the landing page held identical across all of them.
Why: If the bundle offer changes at the same time as the creative, a winning ad and a lucky discount look identical in the results. Freezing everything else is what let us say with confidence that the mess, not a price cut, was doing the work.
Read hook, then hold, then click, then cost
We judged each creative in that order and wrote down exactly which step it failed at, rather than recording a single pass or fail on the ad as a whole.
Why: The eco-friendly creative and the influencer creative both looked like plain losses in a cost-per-purchase column, but they failed at different steps and for different reasons — one never earned attention, the other earned attention and lost trust. That distinction is what the next production round needed.
Promote the winners into a separate scaling campaign
Winners moved out of the test and into a scaling campaign that never gets restructured, while the testing campaign kept its own ring-fenced share of the budget.
Why: The co-op comparison and the refill-maths creatives both needed the retargeting audience to have already seen the stovetop demo. Keeping the winners in a stable campaign structure is what let that sequence build instead of resetting every time a new test launched.
Plan the season four weeks ahead
Six to eight fresh creatives every two weeks, with the pre-Ramadan deep-clean angle and the dust-storm angle shot and queued a month before either season starts.
Why: Both seasonal spikes are predictable to the week in this category. The brand had always shot the Ramadan campaign the week it started, which meant it launched into an auction that competitors had already been warming up for a month.
What changed
The headline result is in the table above, and the shape of it matters more than the size: the cost per purchase came down while spend went up, which is the only version of that number worth reporting. It moved because the share of spend sitting on creatives below the target cost went from a small minority to a clear majority, driven by three of seven test creatives rather than one lucky ad.
The more useful finding was about what actually sells a household cleaner in Kuwait. The three clearest winners each answered a specific, unglamorous worry — is it as strong as it looks, is it safe around my kid, does the refill really pay off — and the two flattest performers were the two that tried to look like something other than a cleaning ad. The brand's instinct had been to make the product look more premium; what actually sold was making the mess look more real.
There was a second effect the founder noticed before we highlighted it in reporting: WhatsApp questions about the co-op comparison and refill pricing dropped once those answers were sitting inside the ads themselves, and the messages that still arrived were closer to a completed order than to a first question.
What we would do next
First, take the refill-maths breakdown onto the product page itself, because right now a shopper only sees that arithmetic if she happens to see the right ad first. That is a conversion job as much as a creative one.
Second, build a dedicated dust-storm response creative that can go live within a day of a warning, since the category's second seasonal spike is weather-triggered rather than calendar-triggered and the brand currently reacts a week too late every time.