SnagRes has logged 61,475,671 availability checks across 283 hard-to-book restaurants in New York City, San Francisco, Los Angeles and Chicago, on Resy, OpenTable, Tock and SevenRooms. This page publishes what that record shows, and the whole dataset is downloadable below. Every figure on every SnagRes page comes from here.
For each restaurant: the share of dates on which a two-top was bookable, the same for a four-top, the hour it releases tables, how far ahead it books, the earliest and latest times that have appeared, the share of openings in the prime dinner window, the weekdays on which it has never opened, and the number of checks behind each of those figures. 39 of the 283 rooms with a full record have never once shown a bookable two-top.
Sorted by the share of checked dates on which a two-top was bookable. Zero means not once.
| Restaurant | Two-top open rate | Four-top | Checks |
|---|---|---|---|
| 4 Charles Prime Rib | 0% | 0% | 645,420 |
| The Ambassadors Clubhouse | 0% | 0% | 613,215 |
| Atomix | 0% | 0% | 4,200 |
| Bavette's Bar & Boeuf - Chicago | 0% | 0% | 52,086 |
| Bungalow | 0% | 0% | 625,043 |
| Buvette | 0% | 0% | 638,846 |
| Carbone | 0% | 10% | 588,600 |
| Cariño | 0% | 0% | 43,266 |
| Cote | 0% | 93% | 578,868 |
| Don Angie | 0% | 0% | 201,735 |
| Enclos | 0% | 3% | 1,571 |
| Feld | 0% | 3% | 1,322 |
| Restaurant | Two-top open rate | Four-top | Checks |
|---|---|---|---|
| A Tí | 100% | 100% | 54,772 |
| Amano Bistro | 100% | 100% | 53,962 |
| Ascione Bistro | 100% | 100% | 53,525 |
| Bar Jabroni | 100% | 100% | 269,972 |
| Bavel | 100% | 100% | 48,774 |
| Beatnik on the River | 100% | 100% | 53,700 |
| Bestia | 100% | 100% | 49,046 |
| Cafe Zoetrope | 100% | 100% | 1,200 |
Each restaurant is rechecked throughout the day, with cadence varying by city, platform and demand. A check records, for a given date and party size, whether a table was bookable and at what times. The open rate is the share of dates in the booking window on which at least one table was bookable at the moment of the check, so it measures what a person refreshing the page would have found. It does not measure how long a table lasted, and a room that opens briefly can read as open on that day.
The per-restaurant dataset is published as CSV at snagres.com/snagres-availability.csv and regenerated with every deploy. A versioned copy with the weekday record and the release schedule is on GitHub at github.com/junnam586/snagres-reservation-dataset, under a CC BY 4.0 license. Cite it as SnagRes Reservation Availability Dataset, with the date. The method, and what the numbers do not measure, is on the methodology page.
Availability is rechecked throughout the day, with cadence varying by city, platform and demand. Figures show the latest completed check. How we measure this.
How we measure · Dataset on GitHub (CC BY 4.0) · The study · Every alert service, compared