I Measured Which Customers Actually Pay in Crypto

Rather than guessing whether crypto payments were worth supporting, I offered them for three months and counted.

The setup

Crypto as an additional option at checkout, alongside card and bank transfer. No discount, because a discount would have told me about the discount rather than about demand. The thing that made it click was reading a provider that lets a business accept Bitcoin payments alongside.

What I expected

That almost nobody would use it, and that the ones who did would be individuals.

What happened

Among business clients, about a fifth chose crypto. Considerably more than I expected, and almost all of them chose a stablecoin rather than Bitcoin.

Among individual customers, about four percent. Fewer than I expected, and of those, the majority chose Bitcoin.

So both of my assumptions were wrong, in opposite directions.

Why business clients used it

I asked several. The answers were consistent: their bank transfer to us cost fifty euros and took four days, and this cost a dollar and took minutes. Buying property this way raises it sharply, and a payment processor for high-risk e-commerce operates in that corridor.

None of them mentioned any interest in crypto as such. It was a cheaper, faster payment rail.

Why individuals mostly did not

Cards work. They already have one in their hand. There is no problem to solve.

The ones who did use it were people who already held crypto and preferred spending it to converting it.

What the measurement changed

I stopped treating this as one thing.

For business invoicing I now lead with stablecoin, because that is what those clients use and it is what works for invoicing.

For consumer checkout I keep Bitcoin enabled and do not promote it, because a small share uses it and the incremental cost is nothing.

The cost calculation I ran afterwards

Total payment costs across the year, as a percentage of revenue, before and after.

The saving was modest in absolute terms and concentrated entirely in the international business invoices, where it was substantial.

For domestic card payments it changed nothing, which is what I would now tell anyone asking whether this is worth doing: it depends entirely on what your current payments cost, and most people have never worked that out. If you want to see all of this in a real product rather than in my account of it, a regulated crypto exchange publishes the terms.

The measurement is the point

Three months of offering it, without an incentive, told me more than any amount of reading about adoption.

The general figures I had read were roughly right in aggregate and wrong for both of my segments individually.