Your Shopper Picked Litecoin and Paid. They Picked Dogecoin and Vanished

A crypto checkout has a step that a card checkout does not. Before any money moves, the shopper has to choose which coin they are paying with, and on which network.

Most merchants treat that screen as plumbing. Our own data says it deserves more attention than that. Across nearly 2.5 million orders where a shopper made that choice, the coin at the top of the table completed at a rate 29 percentage points higher than the coin at the bottom.

Here is the table, the test we ran to make sure a handful of big merchants were not producing it, and then the part we cannot explain.

Want every coin in this table on one checkout? Open a CoinGate account.

What we measured

A CoinGate order runs on two clocks. On the first, the shopper is looking at a list of currencies and has not picked anything yet. Once they choose a currency and a network, the order moves to pending, and from that moment they have 20 minutes to send the payment.

This post only looks at orders that reached pending. The shopper had already decided. What happens next is them either sending the payment inside the window or not.

That matters, because it strips out the part of checkout abandonment that has nothing to do with the coin. Nobody in this dataset wandered off before choosing. It also leaves aside payments that did arrive, only for the wrong amount, which are a different problem and one we took apart separately.

Completion by asset, once the shopper has chosen

A 29-point spread between the top and the bottom. A shopper who picks Litecoin finishes about three times in four. A shopper who picks Dogecoin finishes slightly less than half the time.

The volumes are not close either. Bitcoin and USDC together carry more than half the orders in this table, and both sit in its bottom half.

The order held through the year, too. Litecoin, Tron and Ethereum took the top three places in both halves of the window, and Dogecoin came last in both.

Is it just a few big merchants?

That is the first thing to check on a table like this, and here it is a fair worry. A small number of large merchants carry a big share of every row, so one merchant’s customers could be pulling an asset up or down on their own.

So we ran a harder test. For every merchant with at least 100 orders on an asset and at least 100 on everything else, we compared that asset’s completion with the same merchant’s other assets. The merchant, the products and the checkout page stay the same, and only the coin changes.

Litecoin survives everything. At 96% of the merchants we could test, it completes better than the rest of that merchant’s checkout, by a median of ten points. Tron holds up as well, just less dramatically.

The bottom of the table moves, though. Part of Dogecoin’s last place in the first table comes down to which merchants carry it. Inside the same merchant it lands alongside BNB and USDC, and it is Solana and POL that fall furthest, around ten and twelve points below the rest of the checkout.

The test is thinner for XRP, Dogecoin and POL, which fewer merchants carry at volume. Read those three rows as a direction rather than a precise figure.

The part we cannot explain

We do not know why. We would rather say so plainly, because the two explanations that come to mind first both fail against our own data.

Payment speed does not explain it. That was the first guess, and Bitcoin fits it neatly, since on-chain it is the slowest network here and it sits in the bottom half. However, we have timed the payment clock on every network, and the timings do not line up with these rows. Solana confirms in about a minute at the median and completes ten points below the rest of the same merchant’s checkout. Litecoin takes nearly six minutes and completes ten points above it. In the first table, Ethereum gets paid faster than Litecoin and still completes eight and a half points lower.

Basket size does not explain it either. Litecoin has the smallest median order of the ten, at 8.74 EUR, and the best completion, which looks like a small-basket effect at first. Then you reach the next three smallest baskets, POL at 10.07 EUR, BNB at 10.57 and Solana at 11.16, and all three sit in the bottom half of the second table.

So this is where we are. A large effect that held across both halves of the year and inside the same merchants, with no established cause. The candidates left are things we cannot see from our side, such as who holds which coin, where they keep it, and which wallet is open on which device when the invoice appears.

A number we can stand behind is worth more than a story we cannot.

What a merchant does with it anyway

You do not need the cause to act on the effect.

• Do not drop an asset over this. Dogecoin at 48% still completed 12,098 orders in these twelve months. An asset your customers want is worth keeping even if half of its orders never get paid.
• Do not read a gap as a fault. If Solana or POL completes well below Litecoin on your own checkout, that is what we see at most merchants too. It does not mean something is broken on your side.
• Treat an expired order as a 20-minute problem. Every expired order above is a shopper who chose a coin and did not complete the payment in time. What helps inside that window, and what happens when a payment turns up late, is covered in why crypto checkouts do not convert.
• Measure your own. Our List Orders endpoint returns the status and pay currency of every order, which is everything you need to rebuild this table for your own checkout. Leave out orders with no pay currency, since those never reached the choice. Our book leans towards hosting, IT services and digital goods, so your mix may look nothing like ours.

Method, and what is not in the table

Every figure comes from CoinGate’s own order data, covering orders created between 1 September 2025 and 31 August 2026 that reached pending, meaning a currency and network had been selected. Completion is the share of those orders that were paid, including the few that were later refunded, since those were paid first. The rest expired, were canceled or were marked invalid.

The tables cover the ten assets shoppers can still pay with that have a full twelve months of data. That rule leaves three out. Bitcoin Cash stopped being payable at the end of June 2026. USDT, which is not on our merchant acceptance list, only carries volume in the first five months of the window. EURC only appears in our order data from June 2026. A partial year cannot be compared fairly with a full one, so none of the three is shown.

The USDC row includes orders paid through Binance Pay, which our order data records as USDC. On-chain USDC on its own completes at 55.23%, within a point of the row.

The merchant test compares each asset with the same merchant’s other assets pooled together, and reports the median across merchants, so no single large account can move it.

Finally, these rates count orders rather than people. A shopper who lets one order expire and then starts a new one counts as one expired order and one completed order, and our data cannot link the two. An order can vanish while the customer does not.

Which coin a shopper picks says more about whether you get paid than it has any right to. Litecoin beats the rest of the checkout at almost every merchant we could test, Solana and POL fall behind at most of them, and neither speed nor basket size lines up with any of it.

Thinking it’s time your customers could pay in the coin they actually finish with? Start with us.