Covering Annual Leave in Small Teams Without Burning Everyone Out

Statutory entitlement in the UK is 5.6 weeks a year, which for someone working five days a week is 28 days. Bank holidays can be counted inside that figure or added on top, depending on the contract, and part-timers get the same 5.6 weeks pro rata, so a three-day-a-week colleague has 16.8 days. A team of six full-timers therefore has 168 days of leave to place into a year that contains roughly 260 working days. Divide one by the other and you get 0.65: on an average working day, two thirds of a person is away. In practice that means somebody is on leave in most weeks of the year, before a single day of sickness or training is added.

The more useful number is how many weeks of the year your "one person off" slot is occupied. If only one person may be away at a time, 168 days is 33.6 weeks of the 52, so the slot is in use 65% of the time. In a team of eight it is 86%. That is the moment most managers realise why leave in small teams feels permanently congested: it is not that people are booking too much, it is that a single slot is nearly full even if leave were spread perfectly evenly, and nobody wants February.

Decide your minimum cover before you decide anything else. If six people staff a counter that needs five to run safely, one person off at a time is the rule and you must fit 33.6 weeks of leave into 52, of which about 13 are school holidays that half the team will want. If the job can run on four, two people can be off together and the pressure halves. Write the minimum down per role, not per team: a shop that needs one keyholder and one first-aider on every shift has a different limit from one that just needs three bodies. Our shift coverage checklist walks through how to set those minimums.

Here is what a well-run summer looks like for a six-person team with a minimum cover of five. The eight weeks run from late July to mid September, which covers the English school summer holiday and the first fortnight of term. The rule is one person away on any given day. Two people share week 8 because their days do not overlap, which is the trick that lets a small team squeeze a little more out of a congested window.

Those eight weeks absorb 39 of the team's 168 leave days, which is 23% of the year's leave in 15% of the year, and the slot is full every single day. Nobody has been refused, but nobody has had a second summer week either, and Ben's fortnight only fits because he asked in January. That is a realistic summer, not a bad one. The remaining 129 days have to land in the other 44 weeks, which is where the calendar starts to feel roomy again and why you should encourage autumn and spring bookings early.

First-come-first-served is the default in most small teams because it needs no policy. It rewards people who plan, and it quietly punishes people whose lives are less predictable: parents waiting for a school term-dates letter, carers, anyone whose partner's rota is published late. It also produces the pattern we see most often, where the same two people book Christmas week on 2 January every year and everyone else has stopped asking.

Rota-based booking flips the priority each year. The simplest version is a request window: requests for April to September must be in by 31 January, requests for October to March by 31 July, and clashes inside the window are resolved by a rule rather than by the timestamp on the email. Outside the windows, first-come applies to whatever is left. In our experience this hybrid keeps the fairness of a rota for the peak weeks without making people wait six months to book a Tuesday off in November.

English state schools are closed for about 13 weeks a year: six in summer, two at Christmas, two at Easter and three one-week half terms. If three of your six people have school-age children and each wants two weeks in summer, that is six person-weeks inside a six-week window, which works with one-at-a-time cover only if nobody else wants August. Say that out loud at the start of the year. A team that knows the summer window holds exactly six weeks of leave will divide it up between themselves far more amicably than a team that finds out by refusal.

Christmas is a separate problem because the demand is for a single fortnight and the cover is usually thinner than the rest of the year. Keep a written record of who worked Christmas Day week and New Year week, and make two commitments: nobody covers both weeks two years running, and whoever works the Christmas week gets first pick of the summer window. The second rule costs nothing and is the single most effective piece of leave policy we have seen in small teams.

A six-person team on 28 days each generates 168 days of absence a year, which is 0.65 of a full-time post. Add typical sickness of four to six days per person and you are close to 0.8. That is the size of relief post you would need to cover every leave day fully, and it explains why a part-time relief worker at three days a week is such a common pattern in small care homes, pharmacies and nurseries: it is not a luxury, it is the leave arithmetic made into a job.

If a dedicated post is not affordable, the alternatives are a bank or casual worker who is trained but not rostered, a part-timer on annualised hours whose weeks flex up in summer and down in February, or a planned overtime budget of roughly 168 days multiplied by your shift length. What does not work is borrowing from people's rest days on continuous patterns; we cover why in the guide to building a 4-on-4-off rota, but the short version is that it destroys the recovery time that makes the pattern bearable.

A minimum cover of five is only meaningful if any five people can actually run the shift. In most small teams they cannot, because two or three critical tasks live with one person each. The way to find out is a skills matrix: one row per person, one column per task that must be covered on every shift, and an honest mark in each cell. "Primary" means they do it routinely; "Backup" means they have done it unsupervised in the last three months; "Training" means they are on the way. A blank is a blank.

Read the columns, not the rows. Keyholding has one primary and two backups, which is fine. Bookings has one primary and two backups, but Ben is still training, so the week Chloe and Ellie are both off is the week the diary stops. The matrix tells you two things at once: which pairs of people must never be on leave together, and which training to fund next. Aim for at least two "Primary" or "Backup" marks in every column before you let the calendar decide.

It will happen, usually for the last week of July or the week before Christmas, and the worst way to resolve it is silently by seniority. Use a short ladder that everyone has seen in advance. First, did anyone have that same week last year? If so, they go to the back. Second, was the request made inside the window? Late requests lose to on-time ones. Third, if the tie survives, rotate priority: keep a list, and whoever won the last contested week drops to the bottom. Fourth, before refusing anyone, offer a split, because a Thursday-to-Wednesday week often satisfies two people whose "same week" was really two overlapping half-weeks.

Whatever the outcome, the person who loses should get something concrete: first refusal on the next contested window, or the guaranteed week of their choice outside the peak. A refusal with a promise attached is remembered very differently from a bare no. The same logic applies to handling shift swaps fairly, where a visible rule beats a manager's judgement every time.

All of the above fits on one page, and it should be on one page rather than in six people's memories. State the entitlement in days or hours so that part-timers and shift workers are not arguing about what a "day" is. State the minimum cover per role and the maximum number away at once. State the request windows and the deadline dates. State the clash ladder. State the notice rules: under the Working Time Regulations a worker should give notice of at least twice the length of the leave, and an employer who wants to refuse must say so at least as many days before the leave as the leave would have lasted. Finally, state what happens to untaken days at year end, because carry-over arguments in March are the direct result of nobody having said in April.

Then publish the calendar itself, not just the policy. When everybody can see the summer slot filling up in February, they book March. When the calendar is only in the manager's head, they book in June and find out it is full. A shared view does more for fairness than any rule, because it turns leave from a negotiation with the manager into a visible, finite resource that the team manages between themselves.