Festival Weekend for Hotels: You Sold Out Months Ago. That Was the Easy Part.

Ask a hotel GM how a festival weekend went and you will usually hear "sold out, 100% occupancy, great weekend."

Ask the same GM how the F&B line looked, what housekeeping overtime cost, and how many folios closed clean, and the answer gets quieter.

Selling the rooms during a major local event is the easy part. Demand does that for you. Everything after the rooms is where the weekend is actually won or lost, and most properties plan almost none of it.

Full does not mean profitable

A festival-weekend house looks different from a normal full house in ways that matter operationally.

The guests arrive and leave at the same time. A normal Saturday spreads arrivals from 2 PM to 10 PM. An event weekend puts most of your check-ins in a two-hour band around the time people want to drop bags and go, and most of your departures on a single morning. Front desk queues, bell service backs up, and your early-arrival requests all land at once.

They are out during the hours your outlets make money. Your restaurant's dinner service is empty because the entire house is at the event. Your kitchen is staffed and idle from 6 PM to 11 PM — and then swamped from 11:30 PM by several hundred people who want food now, at exactly the hour you normally close.

They are harder to turn. Late nights mean late checkouts, requested or not. Rooms that a normal Sunday releases by 10 AM come free at 1 PM, and you have a full arrival list waiting. Housekeeping is being asked to do a full-house turn in half the window.

They spend differently. Less on wine and starters, more on late-night food, water, hangover breakfast and minibar. Your standard pars are calibrated for the wrong basket entirely.

The five things worth planning

1. Decide your late-night F&B position before the weekend, not during it

This is the single biggest revenue decision and most properties make it by accident at 11 PM on the night.

Your options are: extend the kitchen, run a reduced late menu, or close and let the delivery apps take it. All three are legitimate. Only one of them is usually chosen deliberately.

If you extend, run a genuinely short late menu — six to ten items that hold and plate fast — and staff it as its own shift rather than as overtime on the dinner shift. If you close, at least decide what the guest sees instead, because a guest who finds nothing at midnight orders delivery for the rest of their stay. That is not one lost ticket; it is the pattern for the whole visit.

The properties that do best here treat the late wall as a service in its own right, with its own menu, its own rota line and its own par levels.

2. Make ordering possible from where the guest actually is

At 12:45 AM your guests are in the lobby, on the terrace, in the corridor, and in bed. They are not going to find the in-room dining directory, decode it, and phone down. They will open an app.

This is where in-room and on-property QR ordering stops being a convenience feature and becomes the difference between capturing the late spend or exporting it. A guest scanning from their room at 1 AM, seeing a short late menu in their own language, and charging it to the room is a ticket you would otherwise have lost outright.

Two details decide whether it works:

Charge to room has to be frictionless but verified. If it takes a phone call to confirm, you have reintroduced the phone call. Server-verified room charge with a PIN or phone match gets the convenience without the fraud risk.

Language. Festival crowds skew heavily international. A menu that answers in the guest's language removes the interaction your night team has least capacity for.

3. Staff housekeeping against the real turn window

Model it honestly. If departures cluster at 11 AM–1 PM and arrivals start at 2 PM, you do not have a turn window — you have a collision.

What actually helps:

• Offer paid late checkout deliberately on the heaviest departure morning. You will get some of it anyway; charging for it lets you predict it and staff around it.
• Split the turn: stayover service in the morning, departures from noon. Do not send attendants down corridors in room order.
• Roster a dedicated runner for linen and amenities so attendants never leave a floor.
• Expect minibar consumption to be several times normal and put restocking on its own pass rather than inside the room turn.

If your scheduling tool projects labour cost against demand, build the housekeeping rota until the coverage warning clears for the turn window specifically, not for the day.

4. Par-up the event basket, not everything

Do not scale every line. Scale the basket this specific crowd buys.

For a music festival crowd that is water, soft drinks, beer, late-night carbs, painkillers and breakfast covers. For a conference it is coffee — several times normal — plus breakfast and bottled water. For a sporting event it is beer, and it is not close.

Minibar and in-room pars deserve their own attention here, because they are the lines that quietly run dry on day one and nobody notices until the third guest complains.

5. Protect the folio

Event weekends produce the worst reconciliation of the year, for predictable reasons: verbal comps, tabs opened and abandoned, charges posted to the wrong room in a rush, and staff too busy to record anything.

The fixes are unglamorous and they work:

• Every charge posted at the point of service, not written on a pad for later
• Room charge verified against the stay rather than by memory of a room number
• A named person responsible for closing the day, with a checklist
• Deposits and pre-authorisations taken at check-in, because chasing a departed festival guest for a disputed charge is a lost cause

Tell the system once, and let it learn

If your platform does demand planning, an event like this should be entered as a date window with a multiplier — and it is worth being clear about what that does and does not do.

It will not infer the surge from the event's name. A festival entered with no multiplier is treated as a completely ordinary weekend. You supply the estimate.

What it does do is correct you afterwards. Once the dates pass, the platform compares actual demand against your baseline and blends the observed ratio into the stored multiplier, so the next occurrence starts from evidence rather than a guess. For an annual event that recurs in your city, the second year is meaningfully better than the first, and the fifth is close to right.

Which makes the argument for recording your guess even if you are not confident in it. An estimate in the system improves; an estimate in your head does not.

From there the multiplier should feed the rota's coverage warnings and the reorder quantities from one place, rather than being re-entered by three different managers with three different numbers.

The Monday after

Two things happen that nobody schedules for.

Your team is exhausted, and your building needs attention — carpets, linen, minibar restocks, and whatever broke. Do not put your deep clean and your stocktake in the same week as the recovery.

And your local trade has stayed away. Regulars avoid the chaos and do not automatically return. The week after an event is when a property finds out whether it traded a weekend of event revenue for a month of soft local business. Something aimed at locals in that recovery week is worth more than it looks on the P&L.

---

Full occupancy during an event is a gift the calendar hands you. The margin is in the late kitchen, the turn window, the right pars and a clean folio — none of which sell themselves.