Dynamic pricing for the restaurant queues tourists travel across the world to stand in.
Some of Tokyo's best restaurants have lines an hour deep, full of people who flew here to eat there, and who'd happily pay to skip the wait. ifasto turns that wait into revenue.
Section 1 · Live demo · paused
This is the actual pricing engine, running live.
Below is a simulated restaurant, but the prices are real. Every number comes from live calls to the same pricing engine a partner venue would use, recalculating every few seconds as the queue changes. Watch what happens as the line grows.
Skip pass activates once the queue reaches 10 parties. Below this, the wait isn't long enough to justify paying for a skip.
The pricing engine is XGBoost-based (L1 wait prediction) feeding a multiplier-based pricing layer (L4) with hard caps on large-party inventory. The system targets a 12–15% conversion band on queued customers, self-regulating in real time. Calibrated against direct field observation from May 2026.
Section 2 · Mechanism
How the price actually gets decided.
The engine looks at four things and combines them into a single skip price. The whole point is restraint: the price is tuned to keep skip-pass buyers at around 12–15% of the queue: enough to be real money for the restaurant, but not so many that the people waiting in the normal line feel like they've been pushed aside.
Why the price can't run away.
Every time someone buys a skip pass, the next one gets a little more expensive. That's deliberate. It means the system slows itself down as demand climbs, instead of selling the whole restaurant to the highest bidders and leaving the regular line furious.
A ramen counter and a sushi temple shouldn't price the same.
The base price is set per restaurant, anchored to what people already spend there. A $96 ramen shop and a $1,920 sushi counter end up with completely different skip prices, automatically. Nobody has to sit there tuning numbers by hand.
Big tables are protected.
Letting a party of six skip means giving up one of your scarcest tables. So by default the engine only allows one large-party skip per service. You can change that number, or turn it off entirely. When the cap's full, the option just disappears.
Section 3 · Restaurant calculator
Run your own numbers.
Put in what your restaurant actually looks like on a busy day. The projection below is built from the same logic the live engine uses. No inflated assumptions, no best-case fantasy.
Your projection
Enter your venue's numbers and click Calculate my upside to see projected weekly, monthly, and annual revenue.
Talk to us about a pilotWhy this exists
I figured this out the slow way, standing in a line in Tokyo.
I'd been waiting far too long, and somewhere in that line I realized the wait was going to cost me half the things I'd planned to do that day. That's when it clicked: people travel across the world for these places, and the thing standing between them and the experience is just … time spent waiting. There had to be a way to fix that.
Arnav
Section 4 · Partnership
The restaurant takes on none of the risk.
I carry all of it: what it cost to build, what it costs to run, and the risk of getting the pricing wrong. The restaurant's side of this is simple: money comes in, nothing goes out.
You pay nothing. I mean nothing.
No setup fee, no monthly fee, no minimum. I cover the cost of building, hosting and running the whole thing. The only line this partnership adds to your books is income: 70% of every skip pass that sells.
It's running by the end of the day.
It's one connection into the POS or reservation system you already use. There's nothing for your staff to learn and nothing new on the floor. The engine runs on my side. None of it lives in your restaurant.
Leave whenever you want.
No annual contract, no notice period. If it's not working for you, one email ends it. I'd rather earn the next month than trap you into it.