Around 12% of sportsbook customers abandon a bet at the confirmation screen when the odds display doesn't match the format they used to build the slip. That figure comes from a 2024 checkout-funnel audit across four regulated markets — the UK, Ontario, New Jersey, and Victoria — covering roughly 2.1 million sessions. It's not a payment failure. It's a formatting mismatch.

The mechanics of the drop-off

Decimal, fractional, and American odds describe the same price. A 2.50 decimal line is 6/4 fractional and +150 American. The maths is identical; the presentation isn't. When a customer browses in fractional and the betslip renders +150, a chunk of them stop.

The audit's breakdown is telling. The 12% overall abandonment splits unevenly:

• Fractional-to-American switches: 15.8% abandonment
• Decimal-to-fractional: 9.4%
• American-to-decimal: 7.1%

Fractional users are the most sensitive. That tracks — 6/4 requires a mental conversion to understand the return, and if the number on the confirm button doesn't match what they were reading thirty seconds earlier, doubt sets in.

Why books still get this wrong

Most operators store odds in a single canonical format and convert at render time. That should work. It often doesn't, because the conversion happens in three places — the market list, the betslip, and the confirmation modal — and each can pull from a different source. A cached market page in fractional and a live betslip in decimal is a common failure mode.

There's also the default problem. A book defaulting to decimal for a UK-facing product will annoy a slice of punters who grew up on fractional, and vice versa for a US-facing product defaulting to American. The 12% figure is the cost of getting that default wrong.

The 2.1 million session sample

Worth noting what the audit did and didn't measure. It tracked sessions where a user added a selection to the betslip and reached the confirm screen. It didn't count users who never got that far, so 12% is a floor, not a ceiling. It also excluded cash-out and in-play, where odds flicker and format matters less than speed.

The markets sampled have different regulatory nudges. The UK and Victoria lean fractional by tradition; New Jersey and Ontario lean American. That mix makes the 12% a reasonable global proxy, though a purely Asian or African sample would likely read differently.

What actually fixes it

The operators with the lowest abandonment in the sample did three things consistently:

1. Locked the format at account creation and never switched it mid-session.
2. Rendered the format in the betslip's confirmation step rather than converting on the fly.
3. Offered a one-tap toggle on the confirm screen itself, so a confused punter can flip to their preferred format without losing the slip.

None of that is technically hard. It's a product decision, not an engineering one.

The question nobody's asking

If a formatting mismatch costs a book 12% of near-complete bets, what does it cost the punter who doesn't abandon — the one who taps confirm on a +150 they misread as 1.50 and stakes ten times what they intended? The audit measured drop-off. It didn't measure regret. That's the number worth chasing next, and it sits on the operator's side of the ledger, not the customer's.