A reported 3x can still lose you money
Start with the number your ads manager shows you. A healthy fashion account in Kuwait reports 2.5x to 4x ROAS on Meta and Snapchat once creative and audiences settle down. Take 1,000 KD of spend at a reported 3x. That is 3,000 KD of tracked revenue, roughly 100 orders at a 30 KD basket. At a 60 percent gross margin you are holding 1,800 KD. Nothing has gone wrong yet, and this is the figure most owners quote to themselves at the end of the month. It is also the last honest-looking number before the deductions start, and in fashion the deductions are brutal.
Now strip it down. Returns run 15 to 30 percent on clothing here because sizing is a guess; at 20 percent you lose 360 KD of that margin plus about 60 KD moving boxes back and forth. If 40 percent of orders are cash on delivery and 15 percent of those fail because nobody answers, the address is wrong, or the mood passed, that is another 120 KD gone. Shipping you subsidise costs 100 KD, KNET and gateway fees 30 KD, packaging 50 KD. You are left near 1,080 KD of contribution against 1,000 KD of ad spend. A reported 3x just bought you 80 KD.
So your real breakeven sits around 2.8x to 3.2x reported, not the 2x everyone repeats. Two things move it faster than any bid change: cut returns with an honest size chart and on-model measurements, and cut failed COD by pushing prepaid. Take 1 KD off the KNET or Tabby order and watch cash on delivery drop within a week. You cannot see any of this inside an ads dashboard, because it needs cost data at the order level, which is why brands that outgrow a link in bio move to a store that reports true margin per order instead of guessing from screenshots.
Ready to start your Shopify store?
Start a free trial and try the platform for yourself.
The first order rarely pays. The second one does.
Everything above assumed the customer buys once. If she buys again, the second order carries no ad cost at all: the same 30 KD basket now returns about 18 KD of gross margin against maybe half a dinar of a WhatsApp broadcast. That is why a first-purchase ROAS of 2x can be perfectly rational for one brand and fatal for another. The variable is repeat rate. Kuwait fashion brands we see land between 15 and 35 percent buying again within 90 days, and the spread is almost entirely about product quality and how the parcel felt when it was opened.
Pull one report before your next budget decision. Take every first-time customer from a single month six months ago and add up everything they have spent since. Divide by the number of customers. If a 28 KD first order became 46 KD over six months, you can afford to acquire at a lower ROAS than a brand still stuck at 30 KD. Most owners have never run this, so they judge every campaign on day-one numbers and switch off the ones that were quietly building a base. Judge prospecting on 90-day value and retargeting on the week.
There is a timing trap on top of that. Kuwait buys hard around Ramadan and the two Eids, then goes quiet from July when half the country travels. A cohort acquired in June looks terrible until October. Do not cut a channel in August on August numbers; hold the comparison to the same month last year. Hala February and the back-to-school run in September are the other two spikes worth planning stock and spend around, and margin in those windows is usually better because you discount less to move the same units.
A big share of your sales never reaches a dashboard
Fashion in Kuwait still closes a lot of business in the DMs. Someone sees a Reel, asks the price on Instagram or WhatsApp, your team sends a KNET link, and the money lands in your account with no campaign attached to it. Snapchat behaves the same way. If you only read platform-reported conversions you will undercount, sometimes badly. Meanwhile the platforms overcount in the other direction: add up what Meta, Snapchat and TikTok each claim in a month and the total will exceed your actual revenue, because all three take credit for the same buyer.
Use blended numbers as the referee. Total revenue for the month divided by total ad spend for the month, every dinar of it, including influencer fees and the agency retainer. That single figure cannot be double-counted and it is the only one your bank account agrees with. Track it weekly. Then add one line to your checkout or your DM script: how did you hear about us. It is crude, only about 60 to 70 percent of people answer, but the pattern it shows across a month is more useful than any attribution model somebody will sell you.
When two channels both claim the same sale, turn one off for two weeks and watch blended revenue. If revenue does not move, that channel was reporting sales it did not create. This is uncomfortable and most agencies will not suggest it, but it is the only test that settles the argument. Run it in a normal month, not Ramadan, not Eid week, not August, and give it a full 14 days so the delivery algorithms and your slower buyers both have time to react before you read the result.