How to Split a Meta Ads Budget Between New Customers, Retargeting and Catalog Ads in Kuwait

Start with about 70 percent of your Meta budget on finding new customers, 15 percent on bringing back recent visitors and 15 percent on catalog ads for specific products. Then move that split every week or two using three signals: frequency, the share of orders from new customers, and blended cost per order. Treat 70/15/15 as a reasoned opening position for a Kuwait store, not an industry benchmark, because no honest benchmark for this exists.

Job one: find new customers. Prospecting means showing ads to people who have never visited your store or bought from you. It is the only job that feeds the other two, because every retargeting audience and every past-customer list starts with someone a prospecting ad reached first. In Kuwait that pool is small and every competitor in your category is bidding for it, so this job needs the largest share and the steadiest supply of new creative.

Job two: bring back visitors. Retargeting means showing ads to people who visited, engaged or added to cart but did not buy. Job three: sell specific products. Catalog ads, which Meta calls Advantage+ catalog ads and previously called dynamic ads, pull items from your product catalogue and show each person the products most relevant to them. The audiences, windows and exclusions for job two are covered in the guide to retargeting in Kuwait, so they are not repeated here.

Why 70/15/15? The two smaller jobs are limited by the size of their audiences, not by how profitable they look, and in a market Kuwait's size a retargeting pool can only absorb so much before the same people see the same ad too often. Worked illustration: on KD 3,000 a month, that is KD 2,100 for new customers, KD 450 for retargeting and KD 450 for catalog ads, or KD 70, 15 and 15 a day. Settle Meta's share of your total with the budget allocator first.

A retargeting audience only refills from new visitors. When an account moves most of its budget into retargeting because that is where the reported return looks best, it stops paying to bring new people in, and the pool it depends on starts to shrink within weeks. The campaign keeps spending, so the same people see more ads, results soften, and the owner concludes that Meta has stopped working in Kuwait.

The reported return was flattering in the first place. Many people in a retargeting pool were already on their way to buying, and the ad that reached them last takes the credit. That is why an account can show a strong retargeting ROAS while total store revenue stays flat, and why the number to trust is total orders against total spend rather than what any single campaign claims for itself.

Kuwait makes this happen faster than larger markets for a simple reason: the audience is small, and the same shoppers are being reached by every store selling the same category. A brand in a large country can live for months on warm traffic from search and marketplaces. A Kuwait store that depends on Instagram for discovery has fewer other sources topping up its pool, so turning prospecting down drains it quickly.

The symptoms are easy to read once you know them. Retargeting reach flat or falling month on month while frequency climbs. A growing share of orders from people who already bought. Good weeks that line up with Ramadan, Eid or White Friday, when interest arrives on its own, and flat weeks in between. If that describes your account, the fix is not a better retargeting ad. It is more budget on job one.

Frequency. Frequency is impressions divided by reach, the average number of times each person saw your ad, which is how Meta's help centre defines it. Check it weekly on your retargeting and catalog ad sets. When it climbs week after week while purchases stay flat, the pool is saturated, so move money back to finding new customers. Use the level at which your own results turned last time, not a threshold borrowed from somewhere else.

Share of new customers. This is first-time buyers divided by all orders in the same period, taken from your store rather than from Ads Manager. Include WhatsApp, phone and cash on delivery orders, which many Kuwait stores take and which never touch the pixel. If the share falls three or four weeks running, the account is recycling the customers it already has, and the split should lean further towards prospecting.

Blended cost per order. This is total Meta spend divided by every order your store took in the period, attributed or not. Worked illustration: KD 3,000 spent and 300 orders is KD 10 an order. If next month the same KD 3,000 brings 280 orders, blended cost rises to about KD 10.71, even if Ads Manager shows a better ROAS. That gap usually means retargeting is taking credit for sales that would have happened anyway.

Move the split in steps of five to ten percentage points, one change at a time, and give each change a week or two before judging it. Changing the split and the total budget in the same week makes it impossible to tell which one moved the result. How far and how fast to raise the total is a separate decision, covered in when to increase your ad budget.

As of September 2026, Meta's help centre calls its automated sales format Advantage+ sales campaigns, the format many advertisers still know as Advantage+ shopping. Pick the Sales objective in Ads Manager and you start in the Advantage+ setup, and Meta says its most advanced optimisation runs when the campaign uses Advantage+ audience, placements and campaign budget. Manual settings are still available, but narrowing the audience or placements switches Advantage+ off for that part of the campaign.

The structural change is that one campaign can now do jobs one and two at once. An Advantage+ sales campaign decides for itself how much to spend on strangers versus people who already know you, and Meta's help centre says the existing customer budget cap, the old setting for limiting spend on past buyers, is no longer available. So a 70/15/15 split is no longer something you type into one campaign. You measure it, then enforce it.

Measure it with audience segments. In Advertising settings you define an engaged audience and your existing customers from custom audiences, and Ads Manager can then break sales campaign results into new audience, engaged audience and existing customers. For a Kuwait store, upload your full customer list, including WhatsApp and cash on delivery buyers, as part of the existing customer definition, or the breakdown will count loyal customers as new ones.

Enforce it with exclusions. Meta's replacement for the cap is excluding your customer audience from the acquisition campaign, which its help centre says does not switch Advantage+ audience off. For a set amount on past buyers, Meta describes two ad sets: one limited to customers with a spending limit, one excluding them. Catalog ads can sit inside the same sales campaign, switched on at campaign level; if they start slipping, read why catalog ads stop selling.

New customers. This creative has to stop a stranger and explain the product in a few seconds, so write the Arabic in Kuwaiti dialect from the brief rather than translating the English, and open with the problem the product solves. Run the Arabic and English versions in the same ad set so Meta can find who responds to which, instead of splitting a small Kuwait audience into two campaigns that are each too small to optimise well.

Returning visitors. These people already know what you sell, so repeating the product pitch wastes the impression. Answer the question that stopped them instead: delivery time to their area, whether you take KNET or cash on delivery, how exchanges work, which size to pick. If your store has separate Arabic and English URLs and enough traffic on each, a website audience built on each path lets you remind people in the language they browsed in.

Catalog ads. The product image and price come from your feed, so the language work sits in the feed and in the text around it. English-only product titles look out of place under Arabic primary text, and a price shown in anything other than dinars reads as a mistake to a Kuwait buyer. Keep the primary text short and suited to the stage, and let the product itself do the selling.

Past customers. This is where Arabic occasion creative earns its place: Ramadan and Eid gifting, back to school, new arrivals in a category they already bought from. Keep it light and tied to what they bought, and keep the budget small, because this is the smallest pool of all. You can start a free Shopify trial to get new versus returning customer data from your own orders, which is what the whole split is measured against.