The room count is a forecast, and you are the one who signed it
A hotel block is a promise in both directions. The hotel promises to hold a set number of rooms at a set rate until a set date. The organizer promises that attendees will book most of them. The second promise is the one with a bill attached: if pickup lands under the threshold in the attrition clause, the organizer pays for the shortfall, in cash, after the event, for rooms nobody slept in.
That makes the room count a financial decision dressed up as a logistics one. It is set at the point in the cycle when the organizer knows least: before registration opens, before the schedule is final, sometimes before the venue is confirmed. The hotel's sales team is happy to help with the estimate, and their incentive runs one way. A bigger block is a bigger guarantee, and a guarantee is worth money to the hotel whether or not the rooms fill.
None of this means blocks are a trap. It means the number deserves the same care as the rate, and that the right question is not how many people are coming. It is how many of them will book this hotel, through this link, before this date. Those are three separate filters, and each one removes more of the audience than organizers expect.
Count travellers, not tickets
The most common sizing mistake is starting from the headline number: tickets sold, guests invited, teams registered. Most of those people do not need a room. Locals sleep at home. Guests within a comfortable drive go back the same night. Some travellers stay with family or friends. The block is only ever drawing from the people who are travelling far enough to need a bed and have nowhere free to put one.
So the first step is to split the audience by where they are coming from. A wedding works from the out-of-town side of the guest list, not the full list. A tournament works from the visiting teams, not the total registration, and then from families per team rather than players. A conference works from the registrants outside commuting range. An event with a past cycle has the best source of all: last time's buyer addresses, which show how much of the audience actually travelled.
Then convert people into rooms. Travellers share. Couples take one room, families take one room, friends split. A roster of players is not a roster of rooms; it is a smaller number of family units, some of which will share a house instead. Counting heads and calling them rooms is how a block ends up far larger than the demand that was ever there to fill it.
From travelling rooms to rooms that book the block
Even an accurate count of travelling households overstates the block, because the block is one option among many and attendees treat it that way. Some have loyalty points with a different chain. Some want a kitchen, a yard for the dog or a house big enough for four families. Some find the block rate is not the cheapest option on their dates. Some never see the link. Every one of those people is housed for the event and none of them counts toward pickup.
Timing removes more. A block closes at its cutoff date, usually weeks before the event, and attendees do not book on the hotel's calendar. Across the events Bookend measures, around 70% of lodging money is committed inside the final 30 days. A large part of the audience makes its lodging decision after the block has already released its rooms, which means it could not have counted toward the block no matter how well the link was promoted.
Nights matter as much as rooms. Contracts are written in room-nights, and the shoulder nights fill differently from the peak. A block that is sized evenly across every night of the event will look healthy on the main night and thin on the first and last, and the attrition calculation may be run on the total. Size each night on its own, from who actually needs to be in town that night, and let the shoulder nights be small.
Why the safe error is to block low
The two ways of being wrong do not cost the same. A block that is too large produces an attrition invoice: a real bill, for a fixed amount, owed by the organizer. A block that is too small produces a sold-out link and some attendees who book elsewhere. The first is a financial loss. The second is an inconvenience, and one that has a good answer, covered further down this page.
That asymmetry is the whole case for sizing to the floor of the forecast rather than the middle of it. Block the rooms you are confident will book through the link before the cutoff, not the rooms you hope will. For a first-time event with no history, that floor is low, and it should be. A block can often be extended if it fills early and the hotel still has rooms; it can almost never be shrunk for free after the dates that allow it have passed.
Organizers resist this because a bigger block feels like better service, and because the hotel offers more concessions as the count rises. Both are real, and neither is worth a guarantee the audience cannot cover. Attendees do not experience the size of the block. They experience whether they found a place to stay. That can be solved without putting the organizer's money behind every room.
Under the hotel's minimum: courtesy blocks, group rates and no block at all
Hotels have a floor of their own. Below a certain number of rooms per night, a contracted block is not worth the sales team's paperwork, and small events often respond by inflating their estimate to clear the line. That is the worst available move: it takes an event too small to need a guarantee and hands it one.
The better options sit below the contract. A courtesy block holds a modest number of rooms until a release date with no financial guarantee at all; unbooked rooms simply go back to the hotel. A group rate is a discount code with no held inventory and no obligation. Both give attendees something to book and give the organizer nothing to owe. For most small and mid-sized events, and for most weddings, one of these is the right default and the contracted block is the exception.
No block at all is also a legitimate answer. An event whose travelling audience is small, scattered or mostly looking for rentals gains little from holding hotel rooms. What those attendees need is a clear place to see what is available near the venue on the event dates. That need exists with or without a block, which is the point of the last section.
The clauses that let the number move after you sign
Because the count is a forecast, the most valuable contract terms are the ones that let it be corrected. The attrition threshold sets how far pickup can fall short before anything is owed; a lower threshold is a wider margin for error on the same block. The basis matters as much: attrition measured cumulatively across all nights forgives a thin shoulder night in a way that night-by-night measurement does not.
A wash clause, sometimes written as review dates, gives the organizer the right to reduce the block by a set portion at set points before the event without penalty. It turns the single early guess into a series of smaller ones, each made with more registration data than the last. An organizer who is unsure of the count should care more about winning a wash clause than about the last few dollars of rate.
Timing of the signature is the other lever. Contract early enough that the hotel still has inventory on the dates, and no earlier than the point where there is some real basis for the number. Where the hotel wants a signature long before registration opens, the answer is a smaller opening block with a written right to add rooms at the group rate if they are available, not a large block signed on optimism.
When the guess was too small, and where everyone else sleeps
A block that sells out early is the good problem, and it still needs handling. Ask the hotel for an extension at the same rate; ask a second property for a courtesy block; and update every page and email that still points at the full link. What should not happen is the default, where the lodging page says sold out and the help stops there. Attendees who hit that message go to a search engine, and the event drops out of every booking that follows.
Those bookings do not go far. In one market Bookend instruments directly, roughly 31% of lodging money after a block sold out still went to properties inside that same block, booked through channels that never counted toward it. The demand was there; the block was simply the wrong container for it. No contract is going to hold the full travelling audience of an event, and sizing as though one could is how attrition bills get written.
The plan that works is a small, safe block for the attendees who book early and want the host hotel, and a live open-market lodging page for everyone else: hotels and rentals near the venue, on the event dates, with real prices. That page carries no guarantee and no cutoff, and it pays. Hotels and hosts pay a standard 15-20% referral commission on the properties in Bookend's panel, and it returns to the event that sent the booking. One peak event weekend Bookend instruments directly produced about 550 room-nights through a single lodging link. With that in place, the block no longer has to be big enough for everyone, so it can be sized to what the audience will reliably book.