Giving farm workers feedback before something goes wrong

On 9 March 2026, the owner of an example corn farm in central Iowa learns that the hired hand has not read the grain bin’s temperature cables for three weeks. Six thousand bushels have heated and sell at $0.80 a bushel below the going price, a loss of $4,800. The hand says nobody told him the readings were his job. The owner has to decide how a worker finds out where he stands before something breaks. The answer is a talk on a date set in advance, held against criteria the worker has in writing, with one line noted the same day. Write those criteria for every job before the next season starts.

Here, criteria means a written list of what counts as done in each task of a worker’s job. The list and the notes from each talk fit on one page per worker, in a notebook or a folder. At the top go the criteria, written before the period you will review starts: one line for each task you will look at, and next to it what counts as done. The worker signs and dates the page when you hand it over. Below go the talks, one line each.

You probably know the other pattern. In the good weeks nobody says anything, so the hired hand or the tractor driver hears from you only when something breaks, and you correct it on the spot or take the job over yourself.

Does a worker know how they are doing if nobody says anything?

Owners tend to think so, and workers do not agree. In a 2018 study of 12 large US dairy farms, 168 workers scored from 1 to 5 how often someone had praised them in the previous 15 days, where 5 meant frequent praise. Owners and managers, guessing how their workers would answer, gave 3.5; the workers gave 2.8. The farms volunteered and were paid to take part, so the numbers describe those farms and not every dairy.

Feedback on the work was scarce too. On 11 of the 12 farms, workers rated how often they heard about their work at less than once a month on average. Their comments show that feedback arriving weekly or monthly was mostly negative. A message like “positive, never; negative, every week” appeared in 22% of the comments offered on that question.

Write down, for each worker, the last day you talked with them about their work when nothing had gone wrong.

What does a worker need in writing before the first talk?

The criteria for their job, in hand before the period you will judge. Edwin Locke, a psychologist, and a colleague reviewed 35 years of research on goals in the journal American Psychologist. Specific goals set at a hard level consistently led to higher performance than urging people to do their best. A do-your-best goal has no external referent, no yardstick outside the person, so each person decides alone what is good enough.

A talk with no written criteria has the same problem, except that you are the one deciding what is good enough, and your answer can change from one day to the next.

The criteria come from the written job description, above all from the result the job is there to deliver. If you have no job descriptions yet, write them first. For each task you will look at, write what counts as done in a way you can both check. On the example farm, the hired hand’s line reads: bin temperature cables read every Monday, each reading logged in the notebook by noon.

Gregorio Billikopf, who wrote the University of California’s textbook on farm labor management, writes that evaluations work best when workers know the evaluation criteria in advance. Rating each person against a standard gives both supervisor and employee a reference point. With the signature at the top of the page, both of you can see that the criteria came before the talks.

Write the criteria for one worker today, one line per task.

How do you talk about the work without talking about the person?

Name the task and leave the person out of it. Avraham Kluger and a colleague pooled 607 comparisons of work with and without feedback on performance. On average, feedback improved performance, but over a third of the time it made performance worse, and it helped less the more it drew attention away from the task and toward the person. The authors add that the effect also depended on task features that are still poorly understood, so nobody can promise that more talk will improve the work.

Billikopf gives the farm version. A vineyard worker may lose heart on hearing he is a bad pruner, while one told he leaves the spurs (the short canes kept at pruning) too long knows what to change. Calling someone lazy or stubborn is likely to yield negative reactions, so he advises talking about the incidents behind that conclusion.

On the example farm, “you are careless with the bin” leaves the hired hand nothing to change. “No reading was logged on Monday, 6 April” gives him the task, the day and the fix.

When the talk shows that the worker does not know how to do the task, more talk will not fix it. In Billikopf’s pruning example the foreman gives more instruction and watches the pruner until it is clear he has understood. A training plan is where you write down who learns what, and by when.

In your next talk, name the task and the day, and leave out any word about the person.

When do the talks happen, and what do you write down?

On dates set before there is a problem, at a rhythm you choose. Richard Stup and a colleague, in a Penn State guide for dairy farms, note that limiting feedback to formal reviews once or twice a year is not enough to help employees improve. Workers with less experience need it more often.

Locke’s review explains why timing matters. For a goal to work, people need summary feedback that reveals progress in relation to their goals. His example is felling 30 trees in a day: nobody can tell if they are on target without knowing how many are down. A fence due on Friday works the same way. The fencer needs to hear on Wednesday how much is up, while there is still time.

No study has measured the right number of talks for a farm. On the example farm, the owner books every second Monday after the readings: 13 April, 27 April and 11 May. A weekly talk that keeps getting moved ends up happening only when something goes wrong.

The talk should come from whoever sees the work. Stup gives the case of a milking supervisor who never visits the night crew and cannot credibly provide them with feedback. If a hired farm manager holds the talks, settle first what your manager decides without asking you.

Write the line the same day. Billikopf notes that supervisors tend to remember events more recent to the evaluation, and that workers, knowing it, try harder as the review gets near. He recommends that critical incidents, moments when someone did especially well or badly, be jotted down as they take place.

Writing also changes what the worker does with the talk. In a review of 138 experiments, most of them about health habits rather than work, Benjamin Harkin and colleagues found that tracking progress toward a goal helped more when the information was physically recorded. On a farm, the closest thing is reading the line aloud at the start of the next talk, so you both begin from what was agreed and not from two memories.

Two things stay off the line: words about the person, such as lazy or careless, and where the worker was hour by hour. Put on the page only what the two of you agreed about the work; it is not for keeping watch on anyone.

The year-end review keeps its place, in Stup’s words, as a summary of the employee’s work over the given period of time. Built from the lines on the page, it holds no surprise, and as Billikopf puts it, there should not be too many surprises when the two of you sit down.

Put the first three talk dates on the calendar before you close this page.

What do you write first, and how long does it take?

Start with one worker and the one task that has cost you the most. Writing the criteria and handing out the pages takes an afternoon if the job descriptions already exist. After that, the only cost is the time of each talk, on the dates you set.

• For each worker, write a page with the criteria for their job: one line per task you will look at, and what counts as done
• Hand each worker their page before the period starts, read the criteria with them, and have them sign and date it
• Put a talk with each worker on the calendar, at a rhythm you can keep and more often for someone new, before any problem comes up
• In each talk, hold the work up against one line of the criteria and name the task, not the person, for what went well and for what has to change
• The same day, write one line on the worker’s page: the date, the line of the criteria, what you saw, what you agreed and when you will check it
• Open the next talk by reading the last line aloud, and at the formal review, if you hold one, read the lines from the period before you judge anything

Following the crew’s work on an ordinary day, with the page open, is a large part of farm management. More guides on the people who work for you are in the people section of this library.