Where the audience number in your grant report comes from

Every arts grant application asks a version of the same question: how many people did you serve? And in most volunteer ensembles the answer is produced the way a fish story is produced — in the week the report is due, by two officers on the phone, saying things like “the Christmas one was packed, call it four hundred.”

Nobody is being dishonest. The number just was not written down on the night, and memory obligingly rounds up. The problem is that the funder's instructions are far more specific than the guess, and the specificity runs in the direction that makes your number smaller.

What the field actually asks for

State arts agencies distributing federal partnership money collect a standard set of descriptive statistics, and the definitions are published. The main one is Actual Individuals Benefiting:

The total number of individuals who were directly involved in the funded activity as artists, non-artist project participants or audience members between the grant or project start and end dates.

NASAA, A Quick and Easy Guide to Data Fields (PDF)

Then the counting instructions, which are where the guess falls apart:

Include actual audience numbers based on paid/free admissions or seats filled. Avoid inflated numbers, and do not double-count repeat attendees.

NASAA, A Quick and Easy Guide to Data Fields (PDF)

Three separate constraints hide in those two sentences.

“Actual… based on paid/free admissions or seats filled.” Not the capacity of the hall. Not the number of programs you printed. A count, taken on the night, of people who were in the room.

“Do not double-count repeat attendees.” If your season is five concerts with a loyal core of 90 people who come to all of them, the sum of your five door counts is not the number of individuals you served. This is the single biggest reason ensemble numbers get quietly reduced by a grants officer.

“Directly involved.” The guide is explicit that figures should cover only those individuals directly affected by or involved in the funded activity, and offers no field at all for indirect reach. So the 4,000 people on your town's e-newsletter list who saw the concert announcement are not participants. Neither, in most framings, are social media impressions.

And there is one instruction almost nobody follows, which is that the honest answer to not knowing is to say so: If actual figures or reliable estimates cannot be secured leave this field blank or enter a ‘-1’ to indicate that data are not available. A blank is a legitimate response. An invented four hundred is not.

The two other numbers you will be asked for

Actual Artists Participating — the artists directly involved in providing the art. For a community band that is your playing roster for the funded activity, and it is the number most likely to be reported wrong in the other direction. Ensembles put down their total membership. The field wants the musicians who actually took part, which after a season of substitutions, leaves of absence and guest players is a different, harder number that only an attendance record can give you.

Youth Benefiting — people under 18, including students and audience members, and it is a subset of the first figure rather than an addition to it. If you run a side-by-side with the high school band, or play a school assembly, this is where that work becomes visible to a funder. Most ensembles who do this work cannot report it, because the student count was never recorded anywhere.

The two-minute habit that fixes all of it

You do not need a data project. You need someone to write down three things before they leave the venue.

• A door count. Ushers with a clicker, or tickets torn and counted, or — the version that works in a church with fixed pews — someone counting filled seats by row during the second piece. Record the method alongside the number, because in two years you will want to know whether 212 was counted or estimated.
• Who played. Attendance taken at the performance, not just at rehearsals. This is what turns Actual Artists Participating from a guess into a figure, and it is also what tells you honestly whether your roster of 74 is really a roster of 74.
• Any youth in the room or on the stage. A separate line. Even approximate, marked as approximate.

Then, once a season, do the de-duplication your funder expects. If you sell named tickets or take reservations, unique attendees fall out of the data. If you do not — and most community ensembles do not — say what you did instead. A sentence like “5,140 admissions across nine performances; we estimate 2,100 unique individuals based on a repeat-attendance survey at the March concert” is worth vastly more to a reviewer than a confident bare number, because it shows you know what the field is asking.

Why this is worth the effort

Grant reviewers read hundreds of these. They can tell the difference between a cultivated number and a remembered one, and the tell is usually that all the figures are round. An ensemble reporting 212, 189 and 341 with a note on method reads as an organisation that runs itself. One reporting 200, 200 and 350 reads as one that does not.

It also compounds. The third year of clean attendance and audience records is what lets you write “audience up 34% over three seasons” in an application, and that is a sentence you can never generate retrospectively.

Rehearsal Letter keeps an audience count against each performance next to the attendance record for the players, so the season's figures exist before the report is due. Requirements differ between funders — check the definitions in your own programme guidelines, since a city arts commission and a federal partnership agreement do not always count the same way.