Planning for Seasonal Peaks and the January Dead Zone

Two weeks in December can carry a restaurant through a quarter. Six weeks in January can quietly undo it.

Most operators know this in their bones and still plan each season the same way: open the calendar when someone asks, roster whoever is available, and hope the covers land where they usually do. That works until the year it does not, and by the time you notice, December is booked at 70% and the rota for the 20th has four people on for a night that needed seven.

The fix comes down to a handful of decisions made earlier than feels necessary. No spreadsheet cult required.

Know the shape of your year, not just the peaks

Every hospitality business has a demand curve, and yours is probably not the one in the trade press.

A city-centre restaurant surrounded by offices lives and dies by the corporate Christmas party, which means its December is front-loaded into the first three weeks and dead by the 23rd. A neighbourhood place with a family crowd sees the opposite: quiet office weeks, then a rush between Christmas and New Year when everyone is home and sick of leftovers. A hotel dining room follows occupancy. A wine bar follows the weather.

Pull last year's covers by day. Not by month, by day. Then do the same for the year before. The pattern you are looking for is not "December is busy", it is "the second Thursday of December is our biggest non-Saturday of the year and we always understaff it".

Two more things worth marking on that curve:

1. The single-night spikes. Valentine's Day, New Year's Eve, Mother's Day where you are, and any local event that fills the town. These behave nothing like a busy Saturday. Demand arrives in a compressed window, guests book earlier, and the whole night is usually two seatings rather than a rolling service.
2. The genuinely dead stretches. Mid-January through early February in most of Europe. The week after a major holiday. Whatever your local equivalent is of "everyone is broke and on a diet".

How far ahead to open bookings

There is no universal answer, but there is a bad default: opening everything twelve months out because the system allows it.

A booking made eleven months in advance is a booking that will be forgotten, and the further out it sits, the higher the no-show risk. It also freezes your inventory before you know your prices, your menu, or whether you will still have that private room.

What tends to work:

Regular service: four to eight weeks. Long enough that planners can plan, short enough that the booking stays real in the guest's head. Reminders close the gap for anything past a fortnight.

Peak single nights: open early and announce it. Valentine's and New Year's Eve reward opening two to three months ahead, because the guests who care are actively looking. The advantage is not just filling seats. It is knowing by mid-January whether Valentine's is tracking ahead or behind last year, while you can still do something about it.

December group bookings: open in September, chase in October. Corporate party bookings are decided by one person in an office who has been told to sort it out. Whoever they can book easily wins. If your December enquiry form is not live by the end of September, you are competing for whatever is left.

The other half of this is the release. Do not quietly open the calendar and wait. Tell your list. A single message to past guests announcing that December or Valentine's bookings are open converts better than almost anything else you will send that quarter, because you are contacting people who already like you at the exact moment they are deciding where to go.

Deposit policy for peak nights

The general case for deposits is weak. The peak-night case is strong, and guests know it.

Nobody is surprised by a deposit for New Year's Eve. Very few are surprised by one for Valentine's, or for a party of ten in December. The expectation is already there, which means the friction cost that makes deposits a bad idea on a normal Tuesday mostly disappears.

Some practical shape for a policy that will not generate arguments:

• Set it per night and per party size, not globally. A deposit on every booking punishes your regulars for the behaviour of December strangers. A deposit on 24 December, 31 December, 14 February, and any table of eight or more is a policy people accept without complaint.
• Take a per-head amount, not a flat fee. Ten euros a head on a table of twelve is meaningful protection. A flat twenty-five euro fee on the same table protects nothing.
• Redeem it against the bill. A deposit that comes off the final total reads as prepayment. A deposit you keep on top reads as a fine.
• Write the cancellation window in plain language and put it in the confirmation. "Cancel by 48 hours before and we refund in full" is a sentence guests can act on. Terms nobody reads produce chargebacks and one-star reviews.
• Have a human override. Someone's flight was cancelled. Refund it, and remember that the goodwill costs you fifty euros and buys you a decade of visits.

For the big set-menu nights, a deposit does something a card hold does not: it commits the guest emotionally. People turn up for things they have already paid for.

Staffing to a forecast, not a hunch

Here is the pattern that costs the most money and gets noticed the least: rostering the same number of people for every Friday in a season where Fridays vary by 40%.

You already have the data to do better. If you know that last year's third Friday in December ran 118 covers and the second Friday ran 74, those are not the same shift and should not have the same rota.

A workable weekly rhythm:

Fourteen days out: look at confirmed covers against the same day last year. Set the provisional rota. Anything running significantly ahead of last year gets an extra body pencilled in now, while people are still free.

Seven days out: recheck. Bookings for a given night typically build in a predictable curve, so if you are at 60% of last year's final number with a week to go and you usually sit at 75% at that point, you have a problem you can still fix with a campaign or a promotion.

Forty-eight hours out: lock it. Confirm the extras, release the ones you do not need, and tell people early enough that they can make other plans. Staff who get cancelled at 4pm remember it.

The forecast also tells you what to buy. December prep is expensive, and the difference between prepping for 100 and prepping for 130 is real money in a month where waste is easy to hide.

The January problem

January is a structural trough, and treating it as a marketing failure to be fixed with heroics burns out your team in the month they most need a breather.

Better framing: January is where you spend the surplus December created, and you spend it on your position rather than on covering costs.

Practical things that work in a trough. Cut your opening days rather than running four half-empty services (two strong nights beat five flat ones, and your team gets a life back). Run the training, the menu development, and the deep clean you never have time for. Push a set-price menu that reads as value without discounting your headline dishes. Book in the private-dining enquiries for spring while you have the time to answer properly. And use the quiet to contact everyone who came to you in December and never came back, because a December guest is a warm contact for a whole year.

That last one is the highest-value thing in the list. Your December book is full of people who visited once, for a party, on someone else's initiative. Most of them will never think of you again unless you give them a reason in January.

The short version

Seasonality is predictable, which makes it the easiest kind of problem in hospitality. Pull last year's covers by day, open peak nights earlier than regular service and tell your list when you do, take a per-head deposit on the four or five nights that genuinely warrant it, and build every rota from confirmed bookings against the same day a year ago. Then plan for January as a season with a purpose rather than a hole you fell into. The restaurants that come out of a winter in good shape are rarely the ones that were busiest in December. They are the ones that knew what December was going to look like in October.