It's the first Sunday in August. The town is heaving. You hired three summer staff in late June, and this morning two of them messaged the group chat to say they're "not coming in today, sorry." One has gone back to her parents in Manchester for the bank holiday. The other has a hangover but has phrased it as "feeling unwell." The third is on shift, but she's only been with you four weeks and still can't close out the till without a manager standing next to her.

You spent five hours yesterday on the floor instead of doing the orders. The pastry case is empty by 11am because nobody pre-portioned the bakes. Margins for August will be down again, and you'll spend September wondering what went wrong.

This is the summer staffing problem. And almost every independent café gets it wrong in roughly the same way.

The 6-week problem

Here is the maths nobody does before they hire.

A new café hire takes 4 to 6 weeks to be net productive. That means they're producing more value per hour than they cost (wage + supervisor time + their mistakes). Before week 4, they're a drag on the business. They cost full wages, they need a senior colleague hovering, and they make the kind of small errors (over-pours, wrong syrups, slow till) that compound across a busy shift.

A 12-week summer hire, on a typical student or seasonal contract, gives you somewhere between 6 and 8 weeks of net productive output. The first 4 to 6 weeks are training and ramp. The last week or two are usually distracted (results day, packing for university, end-of-contract energy).

So the question is not "what does this person cost per hour." The question is "what does this person cost per productive hour, across the whole contract."

If you don't compress the training, the maths breaks. You end up paying nominal £11 to £12 an hour for output that, costed properly, runs at £18 to £22 in the first month. You can absorb that if the back end of the contract is strong. You cannot absorb it if they leave at week 7.

Who actually applies in May and June

Summer applicants fall into three groups, and they are not interchangeable.

Students (school, sixth form, university). Available roughly mid-June to mid-September. Reliable for the middle of summer. Usually disappear around results day or move-in week. Cheaper on NMW if under 21. Generally keen, sometimes brilliant, occasionally vanish without notice when their friends book a last-minute Greece trip.

Returning seasonals. People who worked for you last summer or in a similar venue. Gold dust. Already trained or near it. Worth paying £1 to £2 above NMW to lock in. If you don't have any of these yet, start building the list now for next year - keep notes on every good summer hire and message them in March.

People between jobs. Adults bridging redundancy, a career change, or a relocation. Usually older, often very capable, but the contract end date matters more here. They will leave the minute a permanent role lands. Manage expectations both ways.

The contract terms, the training plan, and the shift design should all flex slightly depending on which group you're hiring from. And if you find a summer hire you'd rather keep for good, taking on an apprentice is a smarter way to turn that seasonal spark into skilled, longer-term staff.

The fixed-term contract: what to put in writing

Every summer hire goes on a fixed-term contract. This is not optional. A handshake "for the summer" gives you neither protection nor clarity, and it makes holiday accrual a nightmare to calculate retrospectively.

The basics:

Term: typically 12 weeks. Pick a start and end date. Write them in.
End date: fixed. The contract ends automatically on that date with no further notice.
Mid-term notice: standard one week either side, unless gross misconduct.
Hours: state minimum guaranteed hours per week (e.g. "minimum 16, up to 35"). Avoid pure zero-hours for under-21s; it makes student loan and NI calculations messy.
Pay rate: state hourly rate, NMW band, and the date their NMW band changes if it does (an 18-year-old who turns 21 mid-contract moves up).
Holiday: state accrual at 12.07% of hours worked (see below).
Probation: two weeks. Not strictly required on a 12-week contract, but useful framing.

Use the same template for every summer hire. Send it through DocuSign or equivalent. Get it signed before day one.

NMW bands and the rate question

National Minimum Wage rates as of writing (2026 rates - check gov.uk before you set yours):

21 and over: £12.21/hr
18 to 20: £10.00/hr
16 to 17 and apprentices: £7.55/hr

The age-based difference is significant. A 17-year-old on the till for 30 hours a week costs roughly £140 a week less than a 22-year-old doing the same job. That gap is real, and the law explicitly allows it.

The question is whether to use it.

My view: use the NMW bands honestly. Pay the legal rate for the age band, but don't grade jobs differently based on age. If the 17-year-old is doing the same work as the 22-year-old, that's fine - they're paid less because the law allows it, not because their work is worth less. If you start them on simpler tasks (clearing, restocking, dishwashing, basic till) and only move them to bar work in week 3 or 4, the lower rate matches the lower output during ramp. That's defensible.

What is not defensible is hiring exclusively under-21s to dodge the adult rate. It's legal but it kills your team chemistry, and your senior staff notice.

The student-specific stuff

Most summer hires will be students, and there are three things that catch out independent operators.

Term-time hours. If a student wants to keep some hours into September or October, check whether their visa or course rules cap weekly hours during term. International students on a Tier 4/Student visa are typically capped at 20 hours during term. Domestic students have no legal cap, but most universities expect under 20.

NI threshold. A student earning under the personal allowance (£12,570) and the NI primary threshold pays no income tax and no employee NI. You still pay employer NI at 13.8% on earnings over £9,100 (the secondary threshold). This is the bit a lot of operators forget. Even a "tax-free" student costs you employer NI on most of their wage.

Student loan. Plan 2 and Plan 5 thresholds (currently £27k+ and £25k+ respectively) mean almost no summer student will hit them. If they're on a postgraduate Plan 3, the threshold is lower (£21k) but still rare for a 12-week contract. Run payroll properly and your software will handle it.

Holiday pay on short contracts

Every hour worked accrues 12.07% holiday pay. On a 12-week contract that comes out to roughly 5 to 6 days of paid leave depending on hours.

You have two clean options:

Pay it weekly as "rolled-up" holiday pay. Add 12.07% to the weekly gross, label it clearly on the payslip as holiday pay. This is now legal again for irregular-hours and part-year workers under the 2024 reforms.
Pay it at the end of the contract. A single lump sum on the final payslip, calculated as 12.07% of total hours worked.

Option 1 is administratively simpler and the staff member sees it weekly. Option 2 keeps weekly take-home cleaner and gives them a small leaving bonus. I prefer Option 1 for under-21s (the weekly visibility helps them understand their pay) and Option 2 for older bridging-job hires (the lump sum lands well).

What you cannot do is ignore it and hope nobody asks. HMRC will eventually.

Pension auto-enrolment

Auto-enrolment kicks in if the staff member earns over £10,000 a year and is aged 22 or over. A 12-week summer hire on £11 to £12 an hour, even at full hours, will struggle to cross £10k of qualifying earnings in that period.

So in practice: most summer hires are below the auto-enrolment threshold and you do not have to enrol them. They have the right to opt in (and you have to contribute if they do, currently 3% employer minimum), but the default is no enrolment.

Document this. A one-line note in the offer letter saying "based on contracted hours, you are not expected to meet the auto-enrolment earnings trigger; you have the right to request enrolment" covers you.

Compressing training to two weeks

This is where the maths actually gets won or lost. The default training plan in most independents is "follow X around for a bit and you'll pick it up." That gives you 6-week ramp time. You need 2-week ramp time.

A 2-week onboarding plan that works:

Days	Activity	Outcome
Day 1-3	Shadow a senior team member every shift. No solo tasks. Watch open, watch service, watch close.	Knows the rhythm of a shift. Can name every drink on the menu.
Day 4-7	Supported solo on simple tasks (till, clearing, simple drinks, restocking, dishwasher). Senior staff visible but not hovering.	Can run till during quiet hours. Makes filter, americano, basic tea orders unsupervised.
Week 2, days 1-3	Full shifts with a planned check-in at start and mid-shift. Take on espresso bar with senior staff nearby.	Can pull and steam. Slow but accurate.
Week 2, days 4-7	Independent shifts, short (4-5 hours). Manager debrief at end of each shift.	Net productive on simple shifts. Still needs support for peak.

This works because the structure is explicit. Both the trainer and the trainee know what "good" looks like at the end of each phase. There's no ambiguity, no "they're sort of getting it."

The cost of this: roughly 30 to 40 hours of senior staff time across two weeks, which is £500 to £700 in supervisor wages depending on your senior rate. That's the training overhead. Bake it into your maths up front.

Shift design for new hires

A common mistake: putting a new summer hire on an 8-hour Saturday because "they need to learn peak." They don't. They need to survive the first three weeks without quitting.

For weeks 1 and 2, schedule short shifts. 4 to 5 hours, mid-week, off-peak where possible. They're learning, and a tired new hire makes more mistakes per hour than a fresh one.

For weeks 3 and 4, ramp to 6-hour shifts including some peak exposure (one weekend day each week, paired with a senior).

From week 5, full shifts. By now they should be net productive and able to handle a busy Saturday with normal team support.

This shift pattern is also a retention tool. New hires who get crushed on day 3 quit on day 7. New hires who feel like they're learning at a sustainable pace stick.

Keeping them through August

You hired them in June. The wobble usually comes in week 5 or 6 - exactly when they're starting to be useful. Reasons: a friend has invited them to a festival, a better-paying job has appeared, they're tired, the novelty has worn off.

A few things that work:

A small bonus at week 6. £100 to £150, framed as "halfway bonus, thanks for sticking with us." This is the cheapest retention spend you'll ever do. Use the pay rise planner to model what a small bump costs you fully loaded.
A larger bonus at end of contract, conditional on completing the term. £200 to £300. State this in the offer letter. It changes behaviour through August.
Flexible scheduling around exams and results day. Ask in week 1 what dates they need off. Put it in the rota. Don't make them ask twice.
Treat them like adults. Brief them on the week's numbers. Show them the menu margin on what they're upselling. People who feel trusted stay longer than people who feel managed.

The September hand-off

Don't think of summer as a one-shot. The best two or three of your summer cohort are exactly who you want for weekend Christmas trade.

In week 10 of the contract, have a conversation with the keepers. Offer a reduced-hours zero-hours or short fixed-term contract from October to early January. Weekends only, plus the December run. They've already trained. They've already proven reliable. Re-engaging them in November costs almost nothing and saves you a Christmas hiring panic.

This is also where the data starts to compound. Year two, you have a list of last summer's keepers. Year three, you have a returning pool. By year five you barely have to advertise.

The real cost per productive hour

Here's the maths, properly costed, for a typical 21+ summer hire on £12.21/hr nominal.

Fully loaded employer cost (NI 13.8% above £9.1k, 12.07% holiday accrual, no pension because under threshold, plus training overhead amortised across the contract): roughly 22 to 25% on top of the nominal wage. Call it £15/hr fully loaded for a working number.

Now divide by productive output across the contract:

Period	Hours worked	Productive output	Real cost per productive hour
Weeks 1-2 (training)	50	~10 productive-equivalent hours	£75/hr
Weeks 3-4 (ramp)	60	~40 productive-equivalent hours	£22/hr
Weeks 5-8 (peak productive)	130	~120 productive-equivalent hours	£16/hr
Weeks 9-12 (steady, slight wind-down)	130	~110 productive-equivalent hours	£18/hr

Blended across the full contract, you're typically at £18 to £22 per productive hour for the first month, dropping to £14 to £16 from week 5 onwards. The contract pays back somewhere around week 7. Anyone who leaves before then has cost you money on net.

This is the number that should drive your hiring decisions, not the headline £12.21.

For a fuller treatment of the headline-vs-real wage gap, see the April 2026 NMW true cost piece. For the broader cost of staff churn, the staff turnover cost write-up is worth a read before you finalise summer headcount.

Where Brikly fits

This is exactly the kind of maths that StaffBrik is built for.

StaffBrik handles fixed-term contracts (start date, end date, fixed-term flag, auto-end), tracks 12.07% holiday accrual per hour worked, calculates fully loaded cost per hour (nominal + NI + holiday + pension + a configurable training overhead), and shows you cost-per-productive-hour as a new hire ramps. So you can see, at a glance, when a summer hire crosses from net cost to net productive.

Pair it with the new hire cost calculator to model a specific contract before you sign it, and the rota planning guide for how to schedule new hires alongside seniors without burning either out.

If you want the wider summer-trading picture, the summer menu shift and iced coffee margin posts cover the demand-side trade-offs that the staffing maths needs to match. And for an overall view of the tooling landscape, the restaurant staff management software round-up maps where StaffBrik sits.

Get the contract right, compress the training, design the shifts properly, and summer goes from margin disaster to your most profitable trading window. Hire reactively in late June and you'll be on the floor again on the August bank holiday wondering where it all went wrong.

Ed O'Brien has run Hunters Cake Company for 17 years across cafés in Witney, Burford, and a bakery in Carterton, Oxfordshire. He's building Brikly - modular tools that give independent café owners the same data the big chains have, without the big chain price tag.