Payment Method Mix Distorts Your Margins - Here's the Fix

The Hidden Cost Layer

Every order has a payment processing cost. Card transactions cost 1.5-2.5%. PayPal charges 2.9% plus a fixed fee. BNPL providers like Klarna and Clearpay charge merchants 3-7% of the transaction value. These aren't small differences when your net margin is 15-25%.

On a £100 order with 20% gross margin, the difference between a credit card payment (£2 cost) and a Klarna payment (£6 cost) reduces your contribution margin from £18 to £14. That's a 22% reduction in available margin - invisible to any bidding strategy that doesn't account for it.

This compounds the COGS accuracy problem that already plagues most POAS implementations. If your cost data is wrong and your payment processing assumptions are wrong, your bidding is doubly disconnected from reality.

BNPL Margin Distortion

BNPL adoption has grown dramatically. In fashion ecommerce, 35-50% of transactions now use Klarna, Clearpay, or similar services. In beauty, it's 25-40%. In home and living, 20-35%. These aren't edge cases - they're a significant portion of your revenue.

The distortion works like this: your POAS calculation assumes a blended payment cost of perhaps 2%. But if 40% of your orders use BNPL at 5%, your real blended rate is 3.2%. On £500k monthly revenue, that's an extra £6,000/month in processing costs your bidding doesn't know about.

BNPL also introduces a refund timing complication. BNPL refunds often take longer to process and may include additional fees. This creates a cash flow gap that further erodes the margin your bidding assumes exists.

Payment Method Economics

Different payment methods have different economics beyond processing fees. BNPL orders tend to have higher AOV (customers spend more when they can spread payments) but also higher return rates (the psychological barrier to returning is lower when you haven't paid in full).

• • Credit/debit card: 1.5-2.5% processing, baseline return rate, immediate settlement
• • PayPal: 2.9% + fixed fee, slightly higher return rate, 1-3 day settlement
• • Klarna/Clearpay: 3-7% processing, 15-25% higher return rate, delayed settlement
• • Apple Pay/Google Pay: Same as card processing, lower cart abandonment, immediate settlement
• • Bank transfer: Lowest processing cost, lowest return rate, but highest checkout friction

When you factor in return rates by payment method, the gap widens further. A BNPL order that's 25% more likely to be returned has an expected margin significantly lower than a card order - yet your bidding treats them identically.

Bidding Implications

You can't control which payment method a customer chooses after clicking your ad. But you can account for the blended payment mix in your margin calculations. This means adjusting your COGS feed to include realistic payment processing costs based on your actual payment method distribution.

For categories with high BNPL adoption (fashion, beauty), this adjustment can reduce your target POAS by 5-10%. That might not sound significant, but at scale it's the difference between profitable and unprofitable campaigns. It's the same principle behind why bundle economics distort bidding - hidden costs that bidding can't see.

The practical fix: calculate your blended payment processing cost by category. Update your margin data monthly as payment method preferences shift. If you're running POAS bidding, this belongs in your cost feed alongside COGS and shipping.

Channel Variation

Payment method preference varies by acquisition channel. Google Shopping traffic tends to have higher BNPL usage than brand search. Social media referrals tend to have the highest BNPL rates. Email and direct traffic have the lowest.

This means your Google Ads campaigns - particularly Shopping and Performance Max - carry structurally higher payment processing costs than your blended average suggests. If you're using blended metrics to set targets, your paid campaigns look more profitable than they actually are. This is one more reason why ROAS looking good doesn't mean profit is up.

Next Steps