Why your board stopped trusting the financial reports
Jeremy Hood, CPA · · 3 min read
A board does not announce that it has stopped trusting the financial reports. There is no vote. What happens instead is that the finance discussion gets shorter, the questions get vaguer, and at some point a board member starts keeping their own spreadsheet.
By the time anyone names the problem, it has usually been developing for a year.
The frustrating part, for an Executive Director who has been pushing for accurate books, is that this rarely has much to do with accuracy. Boards lose confidence in reports that are technically correct. Here is how it actually happens.
The numbers change between distribution and the meeting
The packet goes out on Thursday. On Tuesday, someone circulates a revised version. Maybe the first one was wrong, or maybe a late entry posted, or maybe the Finance Director noticed something and fixed it quietly.
Whatever the reason, the board learns that the number in their hand is provisional. Once that lesson is learned it does not unlearn, and every future figure gets a small mental discount.
The reports disagree with each other
The development report says the organization raised eight hundred thousand. The statement of activities says contribution revenue was six hundred and forty. Both are defensible. One is counting pledges, the other is recognizing revenue, and there is a restriction timing difference underneath.
But nobody reconciles them before the packet goes out, so a board member finds the gap in the meeting and asks about it. The honest answer takes four minutes and involves the word “conditional.” What the board hears is that the finance team cannot explain its own numbers.
This single pattern does more damage than any other. It is also the most preventable.
Variances appear without explanation
A line is forty percent over budget. The report shows the variance in a column. It does not say why.
So the board asks, and the answer arrives verbally, in the meeting, assembled on the spot. Sometimes it is a good answer. It still reads as unprepared, and it puts the Executive Director in the position of defending a number rather than discussing a decision.
A variance column is data. A variance explanation is reporting. Boards can tell the difference.
The format changes
A new treasurer wants a different layout. The accounting system gets reconfigured. Someone builds a nicer template. Each change is reasonable in isolation.
The effect is that the board cannot compare this quarter to last quarter without re-learning where everything lives. Trend is the main thing a board is qualified to assess, and changing the format removes their ability to assess it.
Everything arrives too late to matter
March financials in June are a historical document. The board reviews them, notes them, and makes decisions on instinct, because the numbers describe a situation that has already changed.
A board that cannot act on the reports eventually stops reading them closely. That is a rational response, not a failure of governance.
What rebuilding it looks like
Trust comes back more slowly than it left, and it comes back through repetition rather than through any single improved report.
- Same format, every single time. Change it once a year at most, and when you do, restate the prior period in the new format so comparison survives.
- Distribute once. If the packet is not ready, send it late rather than sending it twice. A late packet is a scheduling problem. A revised packet is a credibility problem.
- Reconcile across departments before distribution. Development, program, and finance should agree on the numbers, or the report should say plainly why they differ.
- Explain variances in the document. One sentence per significant line, written before anyone asks.
- Publish the close date and hold it. A board that knows financials arrive by the fifteenth can build its calendar around them. That predictability is worth more than an extra week of polish.
None of this requires better bookkeeping. It requires someone whose job is the finance function as a whole, rather than the transactions inside it. That distinction is usually what is actually missing.