How to run an executive director performance review

Most boards do not skip the executive director performance review because they think it is unnecessary. They skip it because the chair is a volunteer with a job, the director seems to be doing fine, and there is no date on the calendar forcing the issue. Two or three years pass. Then something goes wrong, or the director asks for a raise, or a board member starts saying quiet things in the parking lot, and the board discovers it has no record of ever having told this person how they were doing.

That is the failure mode. Almost everything below exists to prevent it.

Who runs an executive director performance review

A small committee runs it. Two or three people, usually the board chair plus the vice chair or the chair of another committee, ideally including someone who was on the board a year ago and will be next year.

Not the whole board. A 12-person committee produces a document written by nobody, and the meeting where 12 volunteers discuss one person's performance in front of each other is a bad meeting for everyone in it, including the people who thought it went fine.

Not the chair alone, either, however capable the chair is. A single evaluator turns the review into one relationship, and if that relationship is warm, the review is warm; if it has cooled, the review reads like an indictment. Two or three people arguing about a rating in a room is exactly the friction you want.

Name the committee at the start of the year, not the month you need it. This is a different exercise from a board looking at its own performance, and boards routinely do one and tell themselves they have done both. If that distinction is fuzzy, the difference between assessing the board and evaluating the executive is worth ten minutes before you start.

Evaluate against goals, not impressions

The review should be substantially decided before anyone gathers opinions, because the standard was set 12 months earlier.

At the start of the fiscal year, the board and the director agree on four to six goals. Real ones, with numbers or dates where numbers or dates apply: close the fiscal year within 3% of budget, hire and onboard the program manager by March, get the new evaluation framework in front of the board by the September meeting. Then the review starts by walking those goals in order and asking what happened.

This does two things that impressions cannot. It makes the conversation about work rather than about personality, which is the difference between a review a director can act on and one they merely survive. And it surfaces the board's own contribution, because roughly a third of the time the goal was missed for reasons that trace back to the board: the search committee never met, the strategic plan was never finished, three people said yes to a fundraising task and one did it.

Goals also let you say something specific in a year that was genuinely difficult. "Revenue was down 11% against a plan built before the county grant was cut, and here is what was done about it" is a real assessment. "Everyone feels good about Maria" is not.

How to gather board input without opening a complaint box

Board members do have relevant observations, and the committee should collect them. The risk is that an open invitation for feedback about a person becomes a place to deposit whatever has been annoying someone since the March meeting.

Two things keep it useful. First, ask structured questions rather than for general comment. Ask each board member to respond to the same four or five prompts tied to the agreed goals and to the director's actual job: board support, financial management, communication, program oversight. Second, tell everyone up front what happens to their answers, which is that the committee reads them, summarizes the themes, and brings a summary to the director. Nobody's individual response gets read aloud, and nobody's name gets attached to a sentence.

The committee's job at that point is editorial. Three people saying the board packet arrives too late is a finding. One person's long paragraph about a decision they lost in 2024 is not, and the committee should be willing to leave it out. If your board struggles to say true things at all, the mechanics of getting honest answers from a board apply here as much as anywhere.

One thing to be careful about: board input is not staff input. Asking staff to evaluate their own supervisor, routed through the board, is a different and much more delicate exercise, and doing it casually inside an executive review is how boards accidentally undermine the person they are reviewing.

The compensation conversation, and why the 990 asks

Keep performance and compensation in the same annual cycle but not in the same conversation. Discuss performance, close it, and take up pay separately, usually a few weeks later, with the committee bringing a recommendation to the full board.

There is a specific reason to do this carefully. Form 990, Part VI asks whether the process for determining the chief executive's compensation included a review and approval by independent persons, comparability data, and contemporaneous documentation of the decision. The IRS instructions for Form 990 spell out what each of those three means. Answering yes when you did none of it is a false statement on a public document, and the 990 is the most-read thing your organization publishes whether or not anyone at your organization reads it.

In practice that means the people setting pay have no financial or family relationship with the director, someone pulls comparable salary data for organizations of similar size, mission, and region, and the minutes record who was in the room, what data they looked at, and what they decided. The director should not be present for the vote. This is not bureaucracy for its own sake; it is the documented process that protects both the director and the board if the number is ever questioned.

Write it down, even when it goes well

The output is a short document, two pages is plenty. What the goals were, what happened against each, what the board is asking for in the coming year, and what the board is committing to do. Signed by the chair and the director, both keeping a copy, one filed with the organization's records.

Boards resist this most strongly in the years when everything is fine, which is exactly when it is cheapest to write. The value of a file with four consecutive annual reviews in it is not visible in year one. It becomes visible the first time there is a serious disagreement, a departure, or an unemployment claim, and at that point it either exists or it does not.

The board instrument asks about both halves of this, separately and on purpose: "The board has a written process for evaluating the chief executive" and "The chief executive received a documented performance review in the past 12 months." Plenty of boards can answer yes to the first and no to the second, and that gap is the finding.

The most common failure is never doing one

Everything above assumes a review happens. By a wide margin the most common problem is not a badly run review, a clumsy question, or an awkward compensation conversation. It is three years of nothing, followed by a board that needs to have a hard conversation and has no written record of ever having raised the issue before.

A director in that position is right to feel ambushed, and a board in that position has weakened its own case before it opens its mouth. If you have not done one in years, do not try to build the perfect process. Put a date on the calendar 60 days out, name two people, write down four goals for the coming year even though the year is already underway, and hold a one-hour conversation. A rough review that happens beats an excellent one that stays theoretical.

If you want the board's own view of how it handles this, the assessment asks the board and the chief executive the same questions and keeps the two answers apart rather than averaging them, which tends to reveal whether the board's confidence in its oversight is shared by the person being overseen. It's free, and there's nothing to buy.

Put next year's review date in the minutes at the same meeting where you set the goals. The calendar is doing most of the work here, and it is the only part nobody has to be brave to do.