Net Revenue of Dues and Benefits Given

Dues collected at a level, minus the retail value of the admission that level's members received free or at a discount. It measures the membership program on its own terms and deliberately excludes food, beverage and full-price ticket purchases, which are reported separately.

Benefit given

For a free admission, the value of the seats at the average price actually paid by other guests for that same performance on that same date. For a 10% member discount, the amount discounted. Performances with no full-price sales fall back to the year's average paid seat price ($49.44 in 2023, $53.41 in 2024, $50.26 in 2025).

Memberships sold

Rows on the membership roster. A member who bought more than one membership in a year counts once per purchase. This is the basis the club confirmed for 2023: 585 Basic, 40 Sponsor, 23 Patron, 29 Platinum.

Members

Distinct people on the roster. Lower than memberships sold because of repeat purchases in the same year — 610 people held 658 memberships in 2025.

Active member

A member with at least one transaction that year. Between 88% and 96% of the roster is active in any given year.

Visit

One transaction line. A member who buys two tables for the same show appears twice.

A night, on the Genre Planner

One performance date in one genre. The club booked 382 to 386 of them a year across 2023 to 2025. Per-night revenue is that performance's ticket sales and show-linked food and beverage.

Did not return

A member on the roster in one year and absent from it the next. It does not distinguish a deliberate cancellation from a lapse in renewal.

Benefit use is read from the transaction records, not the discount extract

The workbook's "Member Disc." tabs turn out to cover a different population from the membership roster — in all three years they share no customer IDs with it at all — and they capture only about an eighth of the A-Train activity present in the full transaction records (roughly 1,000 rows a year against roughly 7,000 to 8,000). Benefit given in this report is therefore calculated from the Etix and point-of-sale tabs filtered to roster members, which is the complete picture. Anyone comparing this report to an earlier draft will find benefit given substantially higher here for that reason.

2023 membership IDs were mapped forward

The 2023 membership roster was not put through the Hive customer-ID correction that the 2024 and 2025 rosters received, so its IDs did not match the 2023 transaction records. They were mapped using the old-to-new ID pairs recorded in the corrected tabs; 624 of 676 resolved directly. Of the resulting 2023 roster, 575 members are traceable in that year's transactions.

Two kinds of food and beverage, only one of them usable

About 19,500 rows a year carry an in-venue amount but no performance, no genre and no customer ID. These are the walk-in bar and restaurant side of the export, and they are not reliable: they total $12.7M in 2023, $5.2M in 2024 and $12.1M in 2025, with an average "check" of $764 to $801 in the outer years, which is far too large to be one table's bill — they look like register or shift totals rather than individual checks. Only 37% of them fall on a night with a show in 2023 and 2025, against 97% in 2024, so even the date field behaves differently year to year.

Food and beverage recorded against a ticketed performance behaves properly: $2.87M in 2023, $3.13M in 2024 and $2.82M in 2025, at an average of $95 to $100 a party. Every food and beverage figure in this report is that show-linked spend. The walk-in rows are excluded throughout and are worth raising with whoever maintains the point-of-sale export. Because those rows carry no customer ID, none of the member figures in this report were ever affected by them.

Valuing a free seat assumes it would otherwise have sold

Counting a member's free admission at full retail value assumes the room would have sold that seat to someone else. On a sold-out night that is right; on a quiet Tuesday it is not. The Scenario Planner exposes this as an adjustable assumption rather than burying it — the default of 100% is the most conservative reading of the benefit, not a claim about the room.

Customer names are never published

The data file behind this report contains Customer IDs and figures only. Guest names are held in a separate file that stays on the board member's own machine; the "open name key" control reads it in the browser and joins the names on screen. Nothing is uploaded and nothing is stored on the server. Closing the browser tab clears it.