Most student hires that fail do not fail because of the student. They fail three weeks before the student arrives, at the moment an owner decides to "get some help" and writes a listing describing a person instead of a piece of work. The most personable applicant gets picked. By week three there is nothing specific for them to do, so they invent something, and it is not useful. The owner concludes that hiring college students is more trouble than it is worth. The conclusion is wrong. The sequence was.
I have hired at this level, watched operators in companies I am involved with do it, and sat on a board at Florida International University's Honors College, which is home to more than three thousand undergraduates. The single thing I would tell any Florida employer is that student talent is abundant and cheap to access, and almost none of the difficulty is on the supply side. The difficulty is on yours.
Scope the deliverable before you scope the person
Write down the output first, in a sentence a stranger could grade. Not "marketing assistant." Something closer to "the twenty questions our customers ask before they call, answered properly, published on the site, and kept current." Not "operations support," but "every recurring vendor contract found, summarised on one page each, with renewal dates in a shared calendar."
The test is simple. Could you tell whether the work was done well without being in the room while it happened? If you could not, you cannot supervise it part-time, and a student working around a class schedule is by definition part-time.
Work that suits students is bounded, documentable, tolerant of odd hours, and fails loudly rather than quietly. Work that does not suit them requires institutional memory, judgement under live pressure, or being the only person who knows something. Give a twenty-year-old the second kind and you have not hired help; you have added a dependency you will have to unwind at the worst possible moment.
This is related to but distinct from running an internship. If you want a structured programme with a cohort, a calendar and a credit relationship, I have written separately about how to run one that is worth everyone's time. What I am describing here is narrower and more common: paying a student to do real work for you during term or over a summer, with no programme around it.
Go to the people who already rank them
Posting on a university job board produces volume, not fit. Mass boards select for students who apply to everything, which is a signal about their time management and nothing about their ability. You can do far better by going to the people on campus who already have an opinion about which students are good.
That means instructors who teach capstone or project courses, because they have watched the same thirty students work for a semester and can tell you who finishes. It means officers of student organisations, who are running real operations with real budgets and no authority. It means honors and scholarship programmes, where the students are unusually legible to the faculty who teach them. What strikes me most about that environment is how willing faculty are to make an introduction when an employer asks a specific question — "who in your class can write?" gets a real answer, while "we're hiring" gets none.
Two practical moves. First, offer to speak to a class before you need anybody; show up with a real problem rather than a career talk, and make the ask afterwards. Second, reply to the cold emails students send you. Students who cold-email a local business are pre-selected for the exact trait you are trying to hire, and most of them never get a reply.
Timing matters more than Florida employers expect. The good students commit their semester before it starts. If you begin recruiting in week two of the term, you are choosing from whoever is left.
Pay cash, publish the rate, and raise it when it is earned
Pay them. Unpaid work filters for students who can afford to work for free, which is not the same filter as talent, and it carries legal exposure that is not worth the saving. Check the current wage rules and, if you are considering an unpaid arrangement tied to course credit, check those rules too before you offer anything.
Set one rate per tier of work and put it in the listing. Hiding the number loses you exactly the candidates who are most organised about their time. Your benchmark is not your full-time wage structure — it is the campus job, the restaurant shift, and the tutoring gig the student would otherwise take, plus the flexibility you can offer that none of those can.
Then raise the rate mid-engagement when it is earned, and say plainly why. A modest increase delivered at the right moment, for a named reason, buys more commitment than a higher starting number ever does. Pay on scheduled hours, agree in advance that exam weeks are light, and hold to it when the week comes.
The real cost is not the wage. It is your attention. If you cannot spare a recurring hour a week, do not hire anyone — the wage will be the cheapest part of the mistake.
Screen with a short paid task, not a conversation
Interviews with students mostly test whether somebody has been coached. Replace the second interview with a paid task of two or three hours that resembles the actual work, with a real brief and one piece of deliberate ambiguity in it.
- Did they ask a clarifying question before starting, or guess?
- Did they send it when they said they would?
- Did they send something finished — or a draft with an apology attached?
Those three signals predict more than any transcript. Pay for the task whether or not you hire them, and tell them what you thought of the work. That costs you ten minutes and earns you a reputation on a campus where students talk to each other constantly.
Supervision is the job you are actually signing up for
Students have spent more than a decade working to deadlines against stated criteria. That is the operating system they run. Giving them a written brief, a due date and a definition of good is not infantilising; it is just using the system that already works.
So: a fixed weekly meeting in the same slot, which you do not move. A named person who answers questions within a day. Two examples of work you consider good and one you consider bad, because taste cannot be transmitted by adjective. And a standing instruction that being stuck for more than an hour is a reason to message, not a failure.
Most owner-operators find this harder than they expect, because it is the first time they have had to externalise what they know. That is the same muscle involved in handing work off in the right order, and a student hire is a cheap place to build it. If you cannot write the brief, the problem was never the labour market.
Keeping the good ones after graduation
Decide early and say it out loud. Owners tend to start thinking about conversion in the spring; students start planning in the autumn. By the time you make your move, the good one has accepted something else and felt no obligation to warn you.
The conversation is short and conditional. "If this project lands and you still want to be here in May, there is a full-time version of this role, and here is roughly what it looks like." Name the conditions. A student who knows the bar will usually clear it.
What makes them stay is not ping-pong and not a counteroffer. It is a title that means something outside your company, work that visibly compounds into a skill, and one person who will take their call in five years. Florida gives you a genuine structural advantage here, for reasons I have set out in an honest accounting of talent, cost and proximity — a graduate can build a life on your offer here in a way that is harder in the cities you are competing with.
And when you cannot offer full-time work, say so in February rather than May, and spend an afternoon helping them land somewhere good. The former students of a small business become its quietest referral network.
The second hire costs less than the first
Everything above is reusable. The scoped brief, the paid trial task, the weekly cadence, the named contact on campus — build those once and the next student you hire starts producing in days rather than weeks. That is the part owners miss when they price a student hire as a wage. What you are buying, the first time, is a hiring process you will use for a decade.
We talk a great deal in Florida about the talent pipeline as something institutions are supposed to build. Some of it is ours to build, one scoped project and one well-supervised semester at a time.