What Comp Tickets Actually Cost a Promoter — And Why Sell-Through Won’t Show It

A comp is a sellable seat you gave away. Most promoters know that much.

What costs them is the second half: every box office number you would use to catch the habit is built in a way that makes the room look better as the comp count goes up. Not one convention — both of the ones live entertainment actually uses, by opposite mechanics.

The gross is the only number on the page that comping doesn’t move.

What a comp actually costs

A comp is priced at face value, and it comes out of what you can sell — not out of what you sold. That distinction is the whole accounting problem.

A comp ticket is admission distributed at no charge: the artist’s guest list, press, sponsors, venue staff, industry. Those seats leave the sellable pool as they are allocated — some at contract, some the week of the show. A 1,000-cap room with 150 comps is an 850-ticket room that is still described by its nominal capacity in every conversation you have about it.

That shifts your break-even without shifting your costs. The guarantee, the rent, the production and the marketing are unchanged. The tickets you have left to cover them are not.

At settlement, the comp count is where the arithmetic gets audited. Paid attendance — what you sold out of the sellable pool — is the figure the backend split runs against, and a tour manager who settles for a living will count the list. Comps handed out past the contractual allowance and never logged show up as a gap between the door count and the paid count, and that gap is a dispute.

Convention, not a benchmark: artist lists commonly run somewhere in the range of ten to thirty names for a club or theater show, with a separate promoter allocation for press, sponsors and industry, scaling up for arenas and festivals. Treat that as a starting posture for a negotiation. There is no published average comp count, and anyone quoting you one is quoting a convention too.

Pollstar subtracts your comps from the room

Pollstar removes comps from the capacity before it calculates your percentage sold — so comping raises your reported Sell-Through.

The policy is explicit. Pollstar’s box office reporting rules instruct submitters not to include “complimentary tickets or production kills on the ‘Tickets Sold’ line,” and then go further: submissions should “subtract complimentary tickets and production kills from the capacity of the venue to give an accurate representation of the tickets available for sale” (Pollstar box office reporting policy). That reduced figure is what Pollstar calls Sellable Capacity, and the submission form calculates percentage sold from it.

Run the same show twice to see what that does.

A 1,000-cap room sells 850 tickets with no comps. Sellable Capacity is 1,000, and the show reports 85% sold. Now the same room sells the same 850 tickets, but 150 seats went out on the guest list. Sellable Capacity is 850. The show reports 100% — a sellout.

Same room. Same 850 buyers. At a $32 average ticket, the same $27,200 gross both times. The comps moved reported Sell-Through fifteen points and moved revenue by zero dollars. (Those figures are illustrative — the mechanic is the sourced part.)

This is not Pollstar being careless. Sellable Capacity is the honest way to report how a show performed against what was actually for sale. But it means the percentage in a box office report is not a measure of how full the room was, and it cannot be read as one.

Broadway counts them the opposite way and still flatters you

The adjacent live business includes comps instead of subtracting them, and the fill number still overstates demand.

The Broadway League publishes the rule in its own season statistics: “Attendance represents paid and complimentary tickets. % capacity reflects the percentage of available seats filled (attendance / capacity). Gross represents the Gross Gross. Average paid admission = total gross / total attendance” (The Broadway League, Broadway Season Statistics in Detail). The League switched from Paid Attendance to Total Attendance with the week ending 5/31/09 and footnotes the change on its grosses page.

Comps land in the numerator there rather than being cut from the denominator. The direction of the error is identical: more comps, higher reported capacity.

It also drags the price. Broadway’s 2025–2026 season reported $1,910,903,835 in gross against 14,577,322 admissions, for an average paid admission of $131.09 — gross divided by a headcount that includes people who paid nothing. Whatever the comp share was, it can only push that figure down: $131.09 is a floor on the average a payer paid, not a measure of it.

Two regimes, two mechanics, one lesson: read the gross and the paid ticket count. Those are the figures a comp cannot flatter.

A comp count that grows late is a demand signal

The industry’s own name for filling a soft room with free tickets is papering the house — so a fat or late comp allocation reads as thin advance demand, and reading it is your own diligence.

If the on-sale is tracking and the room is going to fill, nobody needs forty extra names four days out. When that request arrives late in the On-Sale Window, or the allocation quietly doubles, the most economical explanation is usually that somebody closer to the show than you are has looked at the count and decided the room needs help looking full.

Treat it as one input among several, not a verdict. Comp requests spike for ordinary reasons — a local radio partner, a label showcase, a hometown date where the guest list was always going to be long. And papering a soft room on purpose is a legitimate play, which is why the signal is in the timing and the delta, not the raw number.

Where it belongs is alongside the rest of the soft-on-sale evidence: velocity against the opening window, the discounting, the shape of the curve. That diagnostic is its own piece — see what a slow on-sale actually means for the loss-control playbook once the curve has already gone soft, and what independent promoters actually make for where comps sit in the wider margin stack.

To be clear about what tooling can and can’t do here: a Promoter Brief reads an artist’s draw and the economics of your deal. It does not count a specific show’s comps and it does not read a guest list. Nobody’s data feed does. The comp read is diligence you run yourself, on a show you are already inside.

The discipline is a number in the deal memo

A comp allowance is a contract term, and the templates most promoters start from leave it open-ended.

The Performance Booking Agreement published by the Texas Music Office — a state-published template, and a realistic starting point for a promoter without a lawyer on retainer — requires the promoter to “fulfill all of Artist / Band / DJ’s reasonable requests in regards to food and beverage needs and guest list privileges at the Event” (Texas Music Office). No cap, no notice period, no price. “Reasonable” is doing all the work, and it is doing it at your expense.

Not every comp is a loss. A press seat that produces coverage can return more than the ticket was worth. An agent or manager in the room on a show that goes well is worth several future conversations. Those are marketing spend with a plausible return, and they deserve a budget line.

Friends-and-family comps are not that. They are margin, handed over because saying no at the door is awkward — and they are the ones nobody logs.

The fix is unglamorous: a number, in the deal memo, split into an artist allocation and a promoter allocation, agreed before the On-Sale Window opens and reconciled at settlement. Anything past it is a decision somebody has to make on the record, which is the entire point.

The takeaway

When you pull Comparable Shows to size a room or price a deal, read the gross and the paid ticket count — not the sellout percentage. The percentage moves when comps move; the gross doesn’t.

Then set your comp allowance as a written number before the on-sale, and reconcile it at settlement. A comp you can’t tie to coverage, a relationship, or a contractual obligation is margin you gave away and didn’t record.