In this article
Every independent kitchen and floor takes on interns, and almost nobody works out what they actually cost. Not because it is complicated, but because the bill has lines that never end up on the same page.
Do the sum and there are four other lines nobody adds to the stipend: the supervision time of an experienced staff member, the weeks before that intern is running at anything like full speed, the placements that stop before they finish, and the question that decides whether any of it pays off — how many of your interns do you actually go on to hire?
The stipend is the only line that appears on paper, which is exactly why it is the only one anybody ever totals. The other four are buried in a rota that was already full, in a break-in period nobody puts a name to, and in dropouts that only register once the next intern has already arrived.
This guide walks through the seven numbers one at a time, with the arithmetic attached. At the bottom you put your own placement into the calculator — stipend, hours, supervision time, ramp-up pace and conversion odds — and get the full cost per placement and per hour worked.
Everything is worked out in your own browser: nothing is sent anywhere and nothing is stored. The figures below are a worked example for one twelve-week placement — your own kitchen, your own rota and your own conversion rate are the real judge.
Why 'free help' is the wrong question
An intern costs you no wages, and that is the only thing most operators ever add up. But 'no wages' is not the same as 'no cost': every hour an intern is on the floor, an experienced staff member is next to them — briefly, on and off, or for the whole shift — and that is time not spent on something else. That is supervision time, and it has an hourly rate even though it never appears on a payslip.
The second distortion is speed. A new hire who already knows the trade runs at pace from day one. An intern does not — and that is not a criticism, it is precisely the point of a placement. Every week between the first shift and the moment that intern can do roughly what a trained worker does, the station is effectively understaffed. That gap is the ramp cost, and it is the largest of the invisible lines.
And then there is the placement that does not work out. A share of every intake stops early — a bad match, the physical demands of the trade, or simply a training course going a different way. EU vocational-training monitoring — Cedefop's own tracking and the completion figures national apprenticeship authorities publish — consistently puts non-completion in the double digits, with a range that varies widely from country to country and scheme to scheme. What survives that determines what one placement really costs you — and whether conversion into a hire ever gets the chance to pay any of it back.
The ultimate guide Restaurant Staff Guide: 6 Steps to Hire, Train & Retain From hiring to retention: the full staffing system, in one guide. Open the guideThe 7 numbers, and what each one actually costs you
They are ordered by how visible they are. The first is the only one written down anywhere; the last two only show up once you line up more than one intake.
1. The stipend is the smallest line, not the whole bill
A hospitality internship stipend tends to be modest, and on a school placement it is often zero. In the worked example below — twelve weeks, thirty hours a week, a €53 weekly stipend — that comes to €636 for the whole placement. That is the figure that ends up on an agreement, and it is also the only figure most operators ever total.
Put that €636 next to what the rest of this article works out — the supervision time and the ramp cost of that same placement — and the stipend is consistently the smallest of the three. Not because it does not matter, but because the other two are roughly the same size and together weigh more than double.
That is the whole point of this article: 'free help' was always the wrong question, because the stipend was never the whole bill. What is left is not whether an intern costs anything, but how much — and no agreement answers that.
2. Supervision time is borrowed from somewhere else on the floor
Every minute an experienced staff member spends on an intern is a minute not spent on the floor, the pass or the till. Budget twelve minutes of supervision per hour worked — answering questions, watching over shoulders, correcting, explaining again — at the hourly rate of whoever is supervising, and for the example above (twelve weeks, thirty hours a week, €19.95 an hour) that comes to €1,436.
That is more than the stipend itself, and it is money that never gets booked anywhere as an 'intern cost' — it disappears into the supervisor's ordinary wage bill, as if those hours were always going to be there. They were not: that staff member was doing something else with that time before.
The calculator below lets you enter both the supervision minutes and the supervisor's hourly rate, precisely because this is the first of the two biggest invisible lines.
3. The ramp weeks cost output, not money on an invoice
An intern starts at zero and climbs toward a trained worker's pace. That is not a hospitality-specific finding — it is the ordinary operations-management result that new workers on an unfamiliar task follow a fairly predictable but non-linear curve: fast gains early, flattening out as they close in on an experienced worker's speed.
Model that curve as a straight line from zero to full pace over the ramp period, and the average shortfall across that period works out to exactly half of full output. For the worked example — a six-week ramp, thirty hours a week, against a €15.75 hourly wage equivalent — that is €1,418 of output that was not there during those weeks.
That figure never appears on an invoice. It sits in tables that turn slower, dishes that leave the pass more slowly, and a colleague who quietly covers without anyone ever giving it a name. Alongside supervision time, it is the other half of the bill the stipend alone never covers.
Worked example: a six-week ramp, thirty hours a week, against a €15.75 hourly wage equivalent. The shaded part is that week's shortfall — what the station is missing in output.
Over six weeks the shortfall falls in a straight line, from 83% in week one to zero in week six. The average shortfall across the whole period works out to exactly half of full output — the exact assumption the calculator below uses for the ramp cost.
4. Not every placement makes it to the end
A share of every intake stops before it finishes, and that is not a failing of your kitchen alone. Cedefop and the national bodies that track completion figures per vocational programme consistently record non-completion in the double digits — the exact range varies widely by country and by scheme, but the order of magnitude is consistent enough to plan around.
A placement that stops early has usually already cost supervision time and part of the ramp cost, without ever producing a fully productive intern in return. That cost does not disappear — it is quietly carried by the placements that do finish, and that is exactly what the graphic below shows.
It is also why 'we take a few interns every year' is not a strategy on its own, unless somebody is tracking how many of them actually complete the placement. Without that number, you are working from half the bill.
Worked example, spread across the placements that do make it to the end. Stipend, supervision and ramp cost are the real cost of one placement; the fourth slice is what the placements that stop early add on top.
At an illustrative dropout rate of 20% — somewhere in the range Cedefop and national training bodies report per country and per scheme — every placement that does finish carries exactly that share of the bill for the ones that did not. The first three slices are the real cost of one placement: stipend, supervision and ramp cost. The fourth slice is not an arbitrary add-on — it is mathematically exactly the dropout rate, whatever amount sits underneath it.
5. Conversion is the only number that ever pays the bill back
Everything above is cost. There is exactly one line that gives something back: the share of interns you go on to hire. Without conversion, a placement is a training subsidy for somebody else's business — worthwhile, perhaps, but not something your kitchen ever gets anything back for.
Take the full cost of one placement (stipend plus supervision plus ramp cost) and divide it by your own conversion rate to get the cost per hire it actually produces. In the worked example — €3,490 total cost, a 30% conversion rate — that comes to €11,633 per hire. At a 60% conversion rate that figure halves; at 10% it nearly triples.
That is exactly why conversion is the one number worth tracking. Not 'how many interns have we had', but 'how many of them still work here now' — the calculator below puts that percentage directly against the rest of the bill.
6. What one converted hire is worth against hiring cold
A hire that comes out of a successful placement starts with something an interview never gives you: twelve weeks of already having seen how someone behaves on a packed Friday night, how they handle a complaint, and whether they fit the rest of the team. Hiring cold misses all of that, and in a tight labour market finding restaurant staff is already no small cost on its own.
There is no single EU-wide figure for what a bad hire or a long vacancy actually costs you — it varies too much by role, by region and by venue to reduce to one number. What you do have is the cost per hire that comes out of a successful placement, a figure you know yourself once you have filled in the calculator above — and that is the number to set against your own hiring cost, not against an average from somebody else's report.
For some kitchens that comparison shows placements are cheaper than hiring cold. For others it shows the opposite — usually because the conversion rate is too low to ever pay back the supervision time and the ramp cost. Both answers are useful; it is the comparison itself that mostly never gets made.
7. Build a repeatable placement plan instead of improvising it every term
Most kitchens treat every new intern as a first attempt: a rota reinvented from scratch, a supervisor who happens to be free that day, and an induction that depends on whoever has time that week. That is exactly why the supervision time and the ramp cost above run so high — improvising always costs more than a plan that has already been thought through.
A training plan that sets out in advance who shows what, in what order and on which shift, shortens the ramp period without costing more supervision time — it saves it, because nobody is improvising something the last intake already had to learn.
On top of that, simply record who you hire and who you do not, intake by intake, with the reason. After two or three intakes you have your own conversion figure instead of an impression, and that is the one number in this whole article you can deliberately move yourself.
Put your own placement beside it
Fill in what your intern receives, how many hours they work, how much supervision that takes and how fast you expect them to reach full pace. The fields are pre-filled with the worked example above so you can see immediately how it reads — overwrite them with your own numbers.
Underneath sits the full cost of one placement, the cost per hour worked, and — at your own conversion rate — the cost per hire it actually produces.
Intern cost scan
Eight fields, and the full bill from stipend to conversion.
—
This is a model, not an accounting standard: the ramp cost assumes a straight-line climb from zero to full speed, and the conversion rate is your own figure, not an EU average. Everything is worked out in your browser; nothing is sent or stored.
Two things to take from it. Supervision time and ramp cost are both costs you are already carrying, whether or not anybody ever adds them up — working them out does not change them, it only makes them visible. And the cost per hire is not fixed: it falls the moment your conversion rate rises, and nothing in this model moves that faster than a better thought-out plan.
Compare the figure you get here with what a vacancy and a hiring round normally cost you. That is the real comparison — not 'free intern' against 'expensive hire', but two costs that both deserve a number.
What you do with it this intake, this year and next year
Nobody fixes four numbers at once. This order does work, because each step makes the next one measurable.
This intake — run the numbers once
- Fill in the calculator above with the figures from your current or most recent intern.
- Write the three lines down separately: stipend, supervision and ramp cost — and, if you know it, whether that intern finished the placement.
- Ask your supervisor how many minutes an hour it really takes; it is almost always more than anyone assumed beforehand.
- Set the final figure next to what a vacancy and a hiring round normally cost you.
This year — measure instead of guessing
- Track, for every intake, how many interns finish the placement and how many you go on to hire.
- Put a fixed training plan in place, so each intake costs less supervision time than the last.
- Ask whoever supervises what would shorten the ramp period — usually a clearer first week, not a longer placement.
- Work out your own conversion rate after two or three intakes, rather than trusting an impression.
Next year — decide with numbers, not habit
- Set your own conversion rate against the cost per hire and decide whether placements are cheaper than hiring cold for your kitchen.
- Adjust how many placements you take per year to what your supervision capacity can actually carry, not to how many schools ask.
- Put this model next to your other staffing costs in your yearly planning — an intern belongs in the same sum as every other line on the payroll.
An intern is an investment, not a favour
'Free help' was never the right description. An intern costs the supervision time of your best people, a ramp-up period in which the station runs understaffed, and — for a share of every intake — a placement that is never finished. That is not a reason to stop taking interns; it is a reason to cost them the way you cost everything else.
The stipend was always the smallest line. Once you count the other two, the question changes: not 'is this intern free', but 'what does the hire it produces actually cost, and is that less than hiring cold'.
Do that arithmetic before your next intake starts, not after. Every figure in the calculator above is something you can ask your own team today — the supervision time, the ramp pace, and the share who stay.