Should you keep the gala? Event ROI with the staff cost counted

The gala invoice says the event netted $180,000. Nobody added the coordinator’s four months of Tuesdays, the development director’s fall calendar, or the week the finance lead spent reconciling auction payments, and none of that sits on the number the board applauds.

Add the staff hours the ROI formula skips

Every event platform computes ROI the same way: gross revenue minus what you paid vendors. That formula has no line for the people who ran the thing. Add one. Fully loaded net equals gross revenue minus direct costs minus staff hours at a loaded hourly rate, and fully loaded cost per dollar raised equals direct costs plus staff cost, divided by gross revenue. Run both numbers before you touch the fundraising strategy decision this event is actually part of, which is whether the line grows, holds, or gets replaced by something that returns more for the same hours.

What a $260k gala nets once staff time counts

A hypothetical human services organization in its fifteenth year of running a fall gala. Every figure below is invented and carries no source.

That much is what the event platform reports: a net of $180k and a cost per dollar raised of $0.31. Now add the people.

Fully loaded net: $260k minus $80k minus $31k equals $149k. Fully loaded cost per dollar raised: $111k divided by $260k equals $0.43, up from the $0.31 the platform showed. Divide the net by the hours and the event returns $256 for every staff hour it consumes. That is the number that belongs next to the ratio, not instead of it.

What events return against every other channel

AFP puts direct mail acquisition at $1.00 to $1.25 spent per dollar raised, with a renewal mailing to those same donors costing about $0.20. It states plainly that a capital campaign returns much more than an annual program, that a new planned giving effort may show zero return for its first few years, and that a special event returns less than a major gifts program. It does not put a number on events, which is the gap a fully loaded worksheet closes.

Read the two gala rows together. The platform’s $0.31 sits between the two mail figures, which is why a board that only sees that number keeps voting for the gala without a fight. The fully loaded $0.43 sits closer to acquisition mail, the line AFP itself says loses money by design. Neither number is wrong. They price different things.

Neither number is wrong. They price different things.

A good ROI for a fundraising event, once staff time is in the number, is one that beats what the same hours would return on your best other line, not a fixed ratio copied from a vendor’s blog post. A gala at $0.43 fully loaded looks worse than a renewal mailing at $0.20 on paper, until you remember the mailing goes to people who already give and the gala can put a first time donor in a room. Compare the two nets, not just the two ratios, and compare against your own next best use of the hours rather than a published rule of thumb.

Of the four revenue lines in the worked example, sponsorships are the only one whose fully loaded cost sits near the mailing figures once the deck and the ask template exist, because a $1,000 tier and a $10,000 tier cost the development director the same handful of thank you calls. The auction costs the most, because every dollar it raises carries solicitation, cataloging, and reconciliation hours behind it. If the question is what the most profitable fundraising event looks like, the honest answer is not a type of event. It is an event whose revenue mix leans toward sponsorship and away from anything that needs a catalog.

Count what a first time ticket buyer is actually worth

A gala’s financial return is not its only return, and the honest way to count the rest is to price only what you can name, not what you can feel good about. Start with new donors. If sixty of this year’s ticket buyers gave to the organization for the first time, the Fundraising Effectiveness Project’s full year 2024 report found new donor retention at 19.4% nationally, against 69.2% for donors who had given before. Apply that rate and roughly twelve of those sixty come back next year without a second ask. That is the honest floor of what a gala buys in new names, not the number on the guest list.

Sponsor relationships and board engagement belong on the same page, priced the same way. Not “goodwill,” which cannot be counted, but named facts: three sponsors who also fund a program grant, or four board members whose only in-person contact with a beneficiary all year happens at this event. Write down what is specific and countable. Leave off what only sounds like it should matter.

Ticket buyers and auction bidders are not the same audience, and the value calculation above treats them that way on purpose. A ticket buyer who never bid on anything gave once, at a fixed price, mostly for the evening. An auction bidder who also became a sponsor the following year is a major gift prospect the event surfaced, and belongs in next year’s portfolio, not next year’s guest list. Sort the first time names into the two groups before you run the retention math, because a list of two hundred ticket buyers and thirty active bidders is not a list of two hundred and thirty equally likely future donors.

Keep, fix, or cut: the three thresholds

Build a second number before you decide anything: what the same staff hours would return somewhere else. For the hypothetical organization above, redirect the 580 hours to a focused push on forty donors who have given $1,000 to $10,000 before, using personal visits and calls instead of an event. Assume, for the worked example only, a 65% response rate at an average gift of $7,500. That raises $195k. Subtract $2k in materials and mileage and the same $31k in staff cost, and the alternative nets $162k, or $279 per staff hour. Major gifts strategy for organizations under $10M covers how to size and run that push once you decide to build one.

Now the decision has three outcomes, and all three use the numbers you just built.

Keep when the fully loaded net is positive and at or above what the same hours return on your next best line, and the event brings in at least ten new donor households a mailing or a major gifts push would not have reached.

Fix when the fully loaded net is positive but below the alternative, and at least one lever below can plausibly close half the gap without adding staff hours. Run the fix for one year before deciding again.

Cut when the fully loaded net is at or below zero, or the event has failed the alternative test three years running with no fix left to try.

The hypothetical gala nets $149k against an alternative that nets $162k, positive but behind by $13k. That is a Fix, not a Cut. The next section is why.

Three fixes before you cancel anything

Raise the sponsorship floor first. Sponsorship dollars carry almost no marginal direct cost or staff time once the ask template exists, so a $1,000 tier and a $5,000 tier cost the development director the same handful of thank you touches. Moving the floor from $2,500 to $5,000 and adding a $15,000 title tier could plausibly move $20k into the sponsorship line without adding one of the 580 hours, more than the $13k gap between the two nets in the worked example.

Cut the auction second. A live and silent auction usually consumes more staff and volunteer hours per dollar than any other line on the event, between item solicitation, cataloging, bidding software, and day of reconciliation. In the worked example, the $55k the auction raised likely consumed 150 to 200 of the associate’s 340 hours. Dropping it to a handful of donated experience packages, or replacing it with a larger paddle raise ask, frees those hours for the sponsorship ask or the next round of donor visits.

Shorten the program third. A tighter formal program raises the average gift per minute of the paddle raise ask and trims the venue, catering, and AV hours that pad the direct cost line without raising revenue. A three hour program that runs to five costs overtime on both the vendor invoice and the staff time sheet.

The board conversation, with the numbers already run

The gala is usually the board’s favorite night of the year, and a board member who has chaired the auction for a decade will hear “fix or cut” as an attack on the event rather than a read of the numbers. Bring the two net dollar figures, not the two ratios, and name which single lever is closing the gap. “We net $149k against $162k from the same hours elsewhere, and raising the sponsorship floor closes that $13k gap” is a sentence a board can act on. “Our cost per dollar raised went up” is a sentence that starts an argument about the caterer.

Bring this once, at the meeting where next year’s event is approved, not as a standing agenda item that reopens the gala every quarter. Put the fix on the table with an owner and a date, the same way any other budget line gets approved, so the conversation closes rather than repeating in the spring.

The hold or drop call for this line, once you have run it, belongs in the annual fundraising plan alongside every other revenue line, not in a special meeting about the gala. And the board’s own role at the event, the tables they sell and the sponsors they call, is exactly the kind of countable commitment board member fundraising expectations that hold turns into a signed line rather than a vague ask. If the board wants the department wide ratio instead of the single event number, how much should a nonprofit spend on fundraising has the benchmark by budget band.

When not to do this

Do not run this test on an event’s first or second year. A new event is mostly acquisition and setup cost, the same pattern AFP describes for a new planned giving program, and it will fail a test built for a mature event every time.

Skip it if the gala is the organization’s only donor facing gathering all year. A one staff shop with no other cultivation touchpoint gets relational value from the event that a spreadsheet cannot price, and the financial test alone will undercount what the night actually does.

Do not let the sponsorship floor fix become a standing excuse. If the fix does not move the net inside one planning cycle, it was a delay dressed as a plan, and the honest next step is the cut threshold, not a second attempt at the same fix.

And do not hand the worksheet to the board as the whole conversation. Building the fully loaded net is staff’s job, done before the meeting. What the board sees is the recommendation and the one sentence that explains it.