Walk-Ins vs Reservations: Finding Your Restaurant's Balance
How to decide what share of your floor to book and what to hold for the door: the trade-offs, the mechanics (book-and-release, pacing, the waitlist), and how to find your ratio from your own numbers.
July 2, 2026 · 7 min read
There is no universal walk-in-to-reservation ratio, and anyone quoting one is guessing. The right split is a function of your room: how discoverable your door is, how long your turns run, how badly a no-show hurts, and whether a visible queue helps or kills you. What generalises is the trade-off structure and the mechanics for tuning it. Here is how to find your number from your own book.
What reservations buy you, and what they cost
- Buy: predictable covers the kitchen can prep for, guest details before arrival (allergies, occasions), a guest book that compounds, and commitment you can protect with card holds.
- Cost: no-show exposure (an unprotected book is a promise guests can break for free), lost flexibility (a fully committed 7:30 cannot absorb the surprise eight-top), and pacing work.
What walk-ins buy you, and what they cost
- Buy: zero no-show risk (they are already standing there), natural pacing (the queue absorbs demand spikes), the marketing of a visibly wanted room, and spontaneity for the neighbourhood.
- Cost: unpredictable volume for the kitchen, weather risk, walked guests when quotes run long, and no captured contact details unless the waitlist takes them.
Four room types, four defaults
| Room | Sensible default | Why |
|---|---|---|
| Destination dining, long turns | Mostly reserved, deposits on peaks | Prep is committed; an empty two-hour slot is unrecoverable |
| Neighbourhood bistro | Reserved prime slots, walk-in bar and patio | Regulars book, the neighbourhood strolls in |
| High-traffic casual, fast turns | Mostly walk-in with a live waitlist | The queue is the pacing engine; turn-and-burn hates empty held tables |
| Bar-forward rooms | Reservations for tables, bar always open | The bar absorbs waits and builds the walk-in habit |
The mechanics for tuning the mix
Book-and-release: publish most of the book ahead, hold a slice back and release it 24-48 hours out (or day-of) for walk-ins, regulars and recovery. Pacing caps: limit covers per 15-minute slot so the kitchen never takes a wave, whatever the source. The live waitlist: the hinge between the two worlds, converting overflow demand into seated covers and rescuing no-showed tables in minutes. And card holds on the reserved side: they shrink the main cost of reservations, which quietly lets you reserve more of the floor. Operators report no-shows going to nearly zero with a card on file, which changes this whole calculation.
Finding your ratio empirically
- Tally four numbers per service for a month: reserved covers seated, no-showed covers, walk-in covers seated, and walk-ins turned away (or quoted and lost).
- Turned-away walk-ins alongside no-showed tables is the signature of over-booking the floor: you held tables for ghosts while feeding nobody. Protect the book (card holds) or shift slots to the door.
- Empty tables with a thin book means the opposite: open more online slots, and check your booking page is actually linked from Google and Instagram.
- Watch by service, not by week: brunch and Tuesday dinner are different restaurants.
- Re-run the tally each season; the right mix moves with weather, tourists and the block.
This is deliberately your data, not industry benchmarks: published walk-in ratios vary too much by concept and city to borrow, and we will not invent an average. A month of honest tallies beats any number in a blog post, including ours.
The door policy that makes any mix work
Whatever the split, the host stand needs one book that sees everything: tonight's reservations, the live waitlist, and the floor. Split systems (reservations in one app, a paper waitlist, tables in the host's head) are where double-seatings and forgotten quotes come from. Quote honestly and slightly high, take a phone number with every waitlist entry, and go on a wait a few minutes before the room is technically full: the buffer is what keeps quotes truthful.
Turns, covers per service and revenue per seat hour, from your seats and turn time.
Run the turnover calculatorWhat the no-show side of the ledger costs you today.
Run the no-show calculatorCommon questions
What percentage of tables should a restaurant hold for walk-ins?
There is no universal percentage worth borrowing. Tally your own seated, no-showed and turned-away covers for a month; the pattern (ghost tables next to a turned-away queue, or empty floor next to a thin book) tells you which way to move the slider.
Should a small restaurant take reservations at all?
If your turns are long or your guests travel to reach you, yes: predictability is worth more than spontaneity. Fast-turn, high-footfall rooms can thrive walk-in-only with a live waitlist. Many small rooms land on reservations for prime slots plus a walk-in bar.
How do restaurants stop reservations from killing walk-in energy?
Hold back a slice of the floor with book-and-release, keep the bar walk-in-only, and protect the reserved book with card holds so no-shows do not waste tables the door could have filled. The waitlist stitches the two together.