A hiring sheet holds the requirements for one job, the questions every candidate answers and one work test, with a score for each. You write it before the job opens and fill it in the same way for every candidate. It is worth an afternoon. A farm manager lost $80,000 in alfalfa because he trusted an employee who claimed to know how to bale. Write the sheet before the first candidate arrives, and add up the scores before anyone says who they liked.
Is a good conversation enough to pick the right person?
No. A 2022 review of hiring studies rated each way of choosing on a scale from 0 to 1. A score of 0 means no better than drawing names from a hat, and 1 means a perfect forecast of how people later work. The review’s Table 3 puts the methods in this order.
The same fixed questions for every candidate, with each answer scored, came first at 0.42. A work test, where you watch the candidate do part of the job, scored 0.33. An open conversation with no fixed questions scored 0.19, and years of experience scored 0.07.
Read the order, not the decimals. The review does not single out farm work. And 53 of the 54 studies of work tests measured people who already held the job, so the 0.33 is less certain for a stranger at the gate.
Experience is also easy to claim. An applicant for a ranch hand job answered every question with “comes with the territory”; a coworker watched him try to saddle a horse and asked if he had ever ridden one. Talk with the candidate to meet the person, and let the questions and the test do the scoring.
What goes in the two columns of requirements?
The first column holds what rules a candidate out, and the second holds what only helps you rank the candidates left. In the example used here, a lettuce grower in California’s Salinas Valley opens a tractor driver job for the March 2027 planting.
Years of experience go in the ranking column, at most, for the reason in the review: they scored 0.07. What the candidate must do in front of you goes in the first column, since experience is easy to claim and a task done in front of you is not. Billikopf, a farm labor adviser with the University of California, also advises weighting each requirement by how much it matters to the job. Two columns are the simplest way to do that; the split is a practical choice, not a number any study measured.
Copy the requirements from the written job description, the page that sets out what the person in each job does and what the work requires. If this job has none, write it first.
Which questions and which work test do you write, and how do you score them?
Write three to five questions that come from the requirements, ask them of every candidate in the same order, and write beside each one what a good answer sounds like. Three to five is a practical size, not a number any study measured. Billikopf’s guide has a model question for a dairy job. Half an hour before the vet arrives, a milker has not shown up, nobody can take the shift and the milk tank’s cooling is failing: what do you do?
The work test is the same for everyone. Write what you will watch, for how long, and the lowest score that still passes. One dairy in the guide built three stations: loading haylage with a front-end loader, backing a mixer wagon over a set course, and moving three cows through a series of fences. It tried them first on employees and family, to agree on how to score each level.
Score every answer and every part of the test from 0 to 3. Billikopf’s scale reads: 0 cannot do the task, 1 is unlikely to, 2 could with some training, 3 can do it correctly right now. Write each score as it happens. Then add them up before you say who you prefer. A meta-analysis of selection studies found a consistent and substantial loss of validity when data were combined holistically, even by experts who know the jobs, compared with combining by a fixed rule.
Does a referral from someone you trust replace the sheet?
No, and many farm jobs are found by referral. In the US National Agricultural Workers Survey for 2021 and 2022, 61% of crop workers found their current job through friends or relatives. Another 30% applied on their own and 7% were recruited by a grower, foreman or labor contractor.
A referral is a good way to find candidates. Stephen Burks and three colleagues studied nine large firms in call centers, trucking and high tech. Referred workers were 10-30% less likely to quit, and referred truckers had a 12% lower chance of a preventable accident. Those firms are outside farming, so take the direction, not the percentages.
The same study also found that workers refer others like themselves, down to how they behave, so unsafe workers refer unsafe workers. A good referred candidate loses nothing by sitting the same questions and the same test, and the test shows you the one who does not measure up. Write the name of whoever referred each candidate on the sheet. A referral from your best tractor driver is then worth more than one from the worker you would not hire again.
What do you write first, and in what order?
Start with the requirements, because the questions and the test come from them. Steps 1 to 3 are written before the job opens, steps 4 and 5 are filled in as candidates come, and step 6 comes after the hire.
With the scores on paper, a bad hire stops looking like bad luck, because you can see which line missed. A trial period is a weak substitute: Billikopf notes that when a person is barely good enough for the job, he is all too often allowed to stay. To work out what each departure costs your farm, see what turnover costs. Once the person says yes, the next step is the first-day list. A job opens again every time someone leaves, and hiring is one of the farm management decisions you make over and over, so the sheet gets used again and improves with each hire.