The midnight collapse: why your best customers never finish

The sprint problem nobody talks about

Flash deals work on pure adrenaline. Someone sees a flight to Barcelona for £29 or a Taylor Swift pit ticket drop, and they have maybe three minutes to decide. The whole appeal of FlashSeat is that spontaneous rush. You're browsing on the Tube. You pull the trigger. You're done.

Except we'd built checkout like everyone else. Five screens. Address fields that auto-populated wrong. A dropdown menu for every region. Buyer protection checkboxes. By the time someone reached the payment screen, they'd either second-guessed themselves or the deal had sold out.

I remember sitting in a customer support message thread at 1am (yes, we monitor those) where someone wrote: "I tried to book for 15 minutes and gave up. Probably for the best." That line stayed with me. They didn't need protecting from themselves. They needed speed.

What we learned by watching people fail

We did something old-fashioned. We watched users try to check out. Not through heat maps or surveys. We asked 40 people to book a flash deal whilst screen-sharing, and we shut up and listened.

The friction wasn't where we expected. Yes, form fields were annoying. But the real problem was decision fatigue. Users were being asked to think about delivery addresses, payment method selection, and terms and conditions when they were in pure impulse mode. They'd already decided to buy. We were making them decide all over again.

One user, a 26-year-old from Manchester, said something that became our North Star: "I know my address. Why are you asking me?" She had bought from FlashSeat three times. We were asking her to re-enter basic information every single time.

Three steps. That's it.

We rebuilt checkout from zero. Not as a five-screen process, but as a three-step wall. Step one: confirm your saved details (address, payment method). If you're a returning user, this is literally "Is this still right?" with a yes button. Step two: deal summary and final price. Step three: payment confirmation.

For new users, the flows were still there, but we sandwiched them into fewer screens. We made address fields smart. Payment methods defaulted to the most common one. We removed every optional field that wasn't legally required.

The real change, though, was speed. The fastest users went from form to confirmation in 34 seconds. Average time dropped from 4 minutes and 18 seconds to 1 minute and 52 seconds.

We shipped it on a Tuesday. I didn't expect much fanfare. It's checkout, not a feature.

What happened at midnight

The first midnight after launch, I watched the 11pm-to-1am window like a hawk. Abandonment dropped from 62% to 31%. Not perfect, but halved.

By week two, it stabilised at around 28-30%. We still lost people, sure. Some were genuine second thoughts. Some were payment declines. Some hit connection issues. But the majority now got through. Users who wanted the deal booked it.

The real signal came when Premium and Pro members started converting faster than free users. That sounds obvious, but it wasn't automatic. Premium members, who get reduced buyer fees and early access to deals, suddenly trusted the process. They'd saved their payment details. They came back. And they finished.

One member wrote to support: "Finally booked a flight at midnight without panicking the checkout wouldn't work." That sentence is worth more to me than any retention metric.

What else changed when we weren't looking

Shorter checkout had a few side effects we didn't predict. Our support tickets dropped by 18%. Turns out most support emails were people confused by our old checkout flow, not by deals themselves. Fewer questions meant faster response times for genuine problems.

We also noticed Premium and Pro conversion rates ticked up. Not because of the three-step process alone, but because reducing friction made the subscription offer feel less risky. If checkout is snappy, the £7.99-a-month saving on buyer fees suddenly feels more real.

Most interesting: we started seeing repeat flash deal purchases increase. Not just from the same user buying once a month. But from the same user buying three or four times a month. Speed bred confidence. Confidence bred habit.

The thing about midnight buys

Flash deals on flights and event tickets don't happen during office hours. They happen when people are mobile, when they're commuting, when they're bored, when they're at 11:47pm scrolling in bed thinking "wouldn't it be mad if I just booked that Glastonbury ticket?" Midnight is when your best customers actually shop. If your checkout can't keep up, you're not just losing sales. You're losing the kind of users who love spontaneous travel and live events. Those are exactly the people we built FlashSeat for.

The three-step checkout isn't clever. It's just respectful of the moment.