The Cost Structure That Drove High-Risk Merchants First

Crypto acceptance in online retail concentrated almost entirely in categories acquirers price punitively. The concentration is explained by arithmetic. It is worth reading this with the service description of a licensed crypto payment processor open, because the sector differences are mostly differences in which parts matter.

The stack for a high-risk merchant

The headline rate, three to six percent.

Dispute fees of twenty to fifty per occurrence, charged regardless of outcome.

The goods lost on each dispute, which is the order value rather than the margin.

A rolling reserve, commonly five to ten percent held for six months. Capital removed from the business, growing as the business grows.

Monitoring programme fees where dispute ratios exceed thresholds.

And termination risk, which for some businesses is existential.

Effective costs of eight to twelve percent of revenue are common.

Against that

Half to one and a half percent, no disputes, no reserve, no monitoring programme, no scheme termination.

The reason is structural: a confirmed payment cannot be reversed, so the apparatus that prices reversal risk does not exist. Companies building payment products hit this first, which is what a fintech payment gateway exists for.

Why the reserve matters more than the fee

A rolling reserve is not a cost, it is working capital removed for six months, and for a growing merchant it grows continuously and is never fully released.

For a business constrained on cash, eliminating it frequently matters more than the fee saving, which is not visible from a rate comparison.

What it does not address

Fulfilment fraud. The payment is safe and the goods are not, and merchants who relax screening substitute one loss for another.

Refunds, which remain and become the merchant’s to manage.

Customer adoption, which is the binding constraint. A cheaper method used by a small share changes a small share of the cost base.

The measurement

A full year of actual figures, expressed as a percentage of revenue.

Then offer crypto as an additional method for a quarter, without a discount, and measure the share. A discount measures the discount.

Those two numbers are the analysis.

Where the logic generalises

Anywhere conventional payments are priced punitively or refused.

The sector list follows the pricing rather than the technology, which is why predicting adoption from industry characteristics works and predicting it from enthusiasm does not. When something stalls, the difference is whether there is a provider you can actually reach or only a ticket queue.