Key takeaways
- No public dataset gives a reliable national booking lead time; your own reservation export is the best source.
- Online bookings arrive around the clock: an OpenTable estimate put 25% between 10 p.m. and 10 a.m.[1].
- Demand is seasonal, from a 2025 low of $75.8 billion in February to $92.8 billion in May[2], and booking windows stretch as rooms fill.
- Track lead time by day, party size, channel and occasion, then schedule from booking pace.
The honest answer: it depends, and you can measure it
Search for how far in advance people book restaurants and you will find confident single numbers, usually from booking platforms describing their own users. They are a weak guide to your room. Lead time varies by night of the week, party size, price point, neighborhood and season, and a platform-wide average blends all of them together. The public research tells you how and when people book. Your own reservation data tells you how far ahead.
What the research does show is that both channels matter. At top US restaurants in a 2012 Cornell study, 49% of reservations were made by telephone and 25% through OpenTable[3]. Online bookings also arrive while you are closed: OpenTable estimated in 2009 that 25% of its reservations were made between 10 p.m. and 10 a.m.[1]. A booking made at midnight for next Friday is part of your lead-time picture, and it never touches the host stand.
Seasons and occasions stretch the window
Demand swings through the year, and lead times move with it. In 2025, not seasonally adjusted, US food services and drinking places sales were lowest in February at $75.8 billion and highest in May at $92.8 billion[2]. When rooms fill, guests book earlier because the easy times are gone, and the share of same-day bookings falls.
Source: US Census Bureau, 2025[2]
Show the data
| Item | Value |
|---|---|
| February 2025 | 75.8 billion |
| May 2025 | 92.8 billion |
Occasions move the window earlier still. National Restaurant Association research describes Mother's Day as one of the most popular special occasions celebrated at restaurants, if not the most popular[4], and spending on evenings out for Valentine's Day 2026 was expected to reach $6.3 billion[5]. On nights like these the book opens weeks ahead and the last prime tables go early. Plan release dates, deposits and staffing for them separately from ordinary weekends.
Measure your own booking window
Export 12 months of reservations
You need two dates for each booking: when it was created and when the party was due. Keep cancelled bookings, flagged, in the same file.
Calculate lead time
Lead time in days is the reservation date minus the date the booking was created. Same-day bookings are day zero.
Bucket the results
Group bookings into same day, 1 to 2 days, 3 to 7 days, 8 to 30 days and more than 30 days.
Split by what matters
Run the buckets separately for each day of the week, for parties of one to four versus five and up, for phone versus online, and for holidays.
Draw your pace curve
For a typical Friday, work out what share of final covers is already booked 14, 7, 3 and 1 day out. That curve is your forecasting tool.
A spreadsheet is enough. The aim is not a perfect model. It is knowing, on a Tuesday afternoon, whether this Friday is running ahead of or behind a normal Friday, early enough to change the schedule.
Example: reading a lead-time table
| Lead time | Example Friday covers | Example Tuesday covers | What it tells you |
|---|---|---|---|
| Same day | 15 in 100 | 45 in 100 | Weeknights are decided on the day, so staff them from same-day trends |
| 1 to 2 days | 20 in 100 | 25 in 100 | Weekend confirmations matter most in this window |
| 3 to 7 days | 30 in 100 | 20 in 100 | Friday is mostly set a week out |
| 8 to 30 days | 25 in 100 | 8 in 100 | Large parties and celebrations |
| More than 30 days | 10 in 100 | 2 in 100 | Holidays, private events and out-of-town guests |
Example: in this fictional restaurant, roughly two-thirds of Friday covers are normally on the books a week out. A Friday with only half that booked seven days ahead is a warning sign worth acting on, perhaps with an email to regulars. A Tuesday with few bookings three days out is normal. The manager schedules a full Friday team when the seven-day pace is on track and keeps one flexible Tuesday shift that is confirmed on the morning.
Turn the window into staffing and release rules
- Schedule from pace: compare covers on the books with your curve at seven days and three days, and adjust shifts before the final schedule goes up.
- Set release dates: open holiday books on a fixed, announced date, so regulars know when to book.
- Hold back some tables: if most weeknight business is same-day, keep tables for walk-ins and late bookers rather than selling everything a month out.
- Confirm the far-out bookings: plans made weeks ahead are the ones most likely to have changed, so make sure reminders reach them. Guest-facing wording is in reservation etiquette.
- Watch the overnight channel: if a large share of online bookings arrive after closing[1], make sure the online book shows true availability at those hours.
Why party size changes the window
Large parties book further ahead almost everywhere, for two reasons. They are harder to seat, so organizers learn to book early, and they are usually tied to a date that is fixed long before, such as a birthday, a graduation or a visiting relative. Tables of two behave the other way: a couple deciding on Thursday where to eat on Friday is a normal pattern, especially for neighborhood restaurants.
Split your lead-time buckets at whatever party size your floor plan treats as large, often six or eight. If large parties on Saturdays book three weeks out while twos book two days out, your release rules can differ: open large-table inventory early, and keep enough two-tops unsold for the short window. Selling every two-top a month ahead to people who might cancel crowds out the couple who would have booked and come on Thursday.
Lead time and changed plans
The further ahead a booking is made, the more time there is for plans to change. That is not a reason to refuse early bookings, but it is a reason to treat them differently. Look at your own data: compare cancellation and no-show rates for bookings made more than 14 days ahead against those made within three days. If the far-out bookings fail more often, add a confirmation step for them, and use the high chairs and family notes habit of repeating the details in every reminder so guests correct them early.
Where your data is thin, start simple. Even three months of exports split into same-day, this week and further out will show the shape of your demand. Add more detail as the months build up, and re-run the analysis each season, because lead times on a patio in June look nothing like lead times in a quiet February.
Phone bookers and the short window
Phone bookers often behave differently from online bookers. In the 2012 Cornell study, about 65% of diners who had not made an online reservation said they wanted to talk to someone to book[6]. Those calls tend to arrive in the afternoon and during service, often for the same week. If nobody answers then, you lose exactly the short-lead bookings that fill weeknights. Track missed calls by hour next to your lead-time buckets, and staff or cover the phone in the hours your own data shows matter.
What to watch each week
- Covers on the books for each of the next 14 days, against your normal curve.
- Same-day bookings as a share of covers, by day of the week.
- Average lead time for large parties, which should be longer than for tables of two.
- Cancellation timing: how many arrive the day before versus the same day.
- How full the holiday books are on fixed dates before each occasion.
Review these once a week with whoever writes the schedule. Lead time is only useful when it changes a decision about shifts, tables or release dates.
Share the pace numbers with the kitchen as well as the floor. Prep lists, ordering and the number of cooks on the line all depend on the same forecast, and a chef who sees on Wednesday that Saturday is running well ahead of normal can order and prep for it instead of finding out at 8 p.m.
Frequently asked questions
How far in advance should a restaurant open reservations?
Base it on your own data. If almost nobody books more than 30 days out except for holidays, a 30-day window with earlier openings for holidays is enough. Very long windows mostly collect plans that later change.
Do people book further ahead for weekends?
In most restaurants, yes, because weekend prime times fill first. Confirm it by splitting your lead-time buckets by day of the week.
Are same-day reservations worth taking?
Usually. On weeknights they can be a large share of covers. Keep enough capacity and staff for them, and make same-day booking easy online and by phone.
References
Cornell University Center for Hospitality Research (Sheryl E. Kimes) (2009). How Restaurant Customers View Online Reservations (Cornell Hospitality Report Vol. 9, No. 5) (opens in a new tab) ↑
US Census Bureau (2025). Monthly Retail Trade: Estimates of Monthly Retail and Food Services Sales by Kind of Business (1992-present) (opens in a new tab) ↑
Cornell University Center for Hospitality Research (Kimes and Kies) (2012). The Role of Multi-Restaurant Reservation Sites in Restaurant Distribution Management (Cornell Hospitality Report, 2012) (opens in a new tab) ↑
National Restaurant Association (2022). Survey finds restaurants at the heart of Mother’s Day celebrations (opens in a new tab) ↑
National Retail Federation (2026). Valentine’s Day Spending Expected to Reach New Records (2026 survey with Prosper Insights & Analytics) (opens in a new tab) ↑
Cornell University Center for Hospitality Research (Kimes and Kies) (2012). The Role of Multi-Restaurant Reservation Sites in Restaurant Distribution Management (Cornell Hospitality Report, 2012) (opens in a new tab) ↑