Deepavali and Year-End Staffing Coverage Without the Overtime Blowout

If you run a Singapore SME, your year-end coverage plan needs to be finalised in the next three weeks — not in the week before Deepavali. The working answer has three parts: publish a written coverage roster for every holiday and leave-heavy week from November through early January by mid-October; agree time-off-in-lieu in writing, before the work happens rather than settling overtime claims after it; and remove enough routine enquiry volume through automation that you need fewer people on duty in the first place. Year-end manpower budgets do not blow out because staff are greedy. They blow out because the roster gets decided late, under pressure, by whoever happens to still be answering the phone.

What does the 2026 year-end calendar actually look like?

Map the pressure points before you roster a single person. The compression this year is worse than most owners realise:

• Deepavali falls on Sunday 8 November 2026, with the public holiday observed on Monday 9 November. That is a long weekend, not a single day off — and for teams with a significant Indian workforce, leave requests will cluster around it on both sides.
• Christmas Day is Friday 25 December 2026 and New Year's Day is Friday 1 January 2027. Two consecutive long weekends with a dead zone between them, where a third of your team is on leave and your customers' finance departments are running skeleton crews.
• Chinese New Year 2027 arrives on 6–7 February. That is close enough that January becomes a backlog-clearing month, not a recovery month. Anything you push from December does not get absorbed — it gets pushed again.

The trap is that leave demand and customer demand peak in the same weeks. Retail, F&B and logistics see obvious December volume. But B2B firms get hit just as hard, because customers chase year-end delivery, year-end invoicing and next-year budget commitments in exactly the fortnight your team wants off.

Why does year-end overtime blow out instead of holding flat?

Because late rostering is expensive rostering. When coverage is decided on the Wednesday before a long weekend, you have no options left except paying whoever is available. Planned coverage gives you choices — swaps, staggered leave, time-off-in-lieu, a half-day on-call shift instead of a full one. Panic coverage gives you one: overtime at premium rates, usually from your most senior and most expensive person, because they are the only one who can handle an escalation unsupervised.

The mechanics matter here. Under Part IV of the Employment Act, overtime entitlements apply to workmen and to non-workmen earning up to the prescribed salary thresholds, paid at 1.5 times the hourly basic rate, with a monthly overtime cap. Work performed on a gazetted public holiday attracts an extra day's salary, or a day off in lieu by mutual agreement. Rest day work is calculated differently again. Confirm the current thresholds, rates and cap on the MOM website before you build your numbers — they are reviewed periodically, and the figure you remember from two years ago may not be the figure you owe.

The part that never appears in the payroll report is the manager time. An owner spending eleven days in December personally covering enquiries because the roster failed is not an overtime line item, but it is the single most expensive labour cost in the business.

Which year-end work can you delete instead of staffing?

Before you roster more hours, cut the volume those hours exist to absorb. Pull your last December's enquiry log and categorise it. In most SMEs, over half of year-end contact volume falls into four buckets that do not require a human at all:

• Order and delivery status. The single largest December call driver. A self-service status link sent automatically at dispatch removes most of it. Every one of these calls answered by a person is a call your team could have spent on a quotation.
• Operating hours and holiday closure questions. Publish your full November–January closure schedule now on your website, Google Business Profile and WhatsApp business profile. Update it once; it answers hundreds of times.
• Repeat orders from existing customers. These follow a known pattern. A templated reorder flow turns a five-minute conversation into a thirty-second confirmation.
• After-hours first contact. This does not need a person — it needs a commitment. An auto-response that states a specific next-business-hour reply time, and is actually honoured, holds an enquiry far better than silence.

The point of automating these is not novelty. It is that every deflected routine contact is a roster hour you do not have to buy at 1.5 times.

What does a non-chaotic coverage roster look like for a small team?

For a team of ten to fifteen, a workable plan fits on one page:

1. Close leave applications for the 8 Nov – 5 Jan window by 15 October. Announce the cut-off now. First-come-first-served, with a stated maximum number of people off per function per day.
2. Define minimum viable coverage per day, not per week. For most SMEs this is one person who can answer enquiries and one who can authorise a quotation or a delivery exception. That is often two people, not eight.
3. Name a backup for every primary. If a role has exactly one qualified person across the whole window, that is key-person risk, and December is when it detonates.
4. Agree time-off-in-lieu in writing before the shift. Specify the date the lieu day will be taken, and take it in January. Undocumented lieu days become disputed overtime claims in February.
5. Set an after-hours response standard and staff to it. A four-hour target on a public holiday is honest and achievable. A vague promise to monitor messages is neither, and it quietly turns everybody's holiday into unpaid on-call.

What should you do before mid-October?

Three things, in order. Publish the leave cut-off date this week. Pull last December's enquiry volume and identify the top two categories you can automate away by 1 November. Then draft the roster, circulate it, and let people trade shifts among themselves — staff will solve a published roster far more cheaply than you can solve an unpublished one.

Do this in September and year-end coverage is a scheduling exercise. Do it in November and it is a bidding war for your own employees' time.

Frequently asked questions

Can I require staff to work on a public holiday?

Generally yes, subject to the employment contract and the Employment Act, but it triggers specific entitlements — an extra day's salary at the basic rate, or a mutually agreed day off in lieu, with overtime on top if the employee works beyond normal hours and falls within Part IV coverage. Agree the arrangement in writing before the day, and confirm the current rates with MOM rather than relying on precedent from a previous year.

Is time-off-in-lieu always cheaper than paying overtime?

In cash terms usually yes, but not always in practice. Lieu days taken in a busy January can create a second coverage gap that you then fill with overtime anyway. The fix is to schedule the lieu day at the same moment you agree to it, in a week you know will be quiet, rather than leaving it as an open balance.

We are too small to automate anything. Where do we start?

Start with the single highest-frequency question your team answers in December, which is almost always order or delivery status. Automating one question does not require a platform migration — it requires a status link, a standard message and a decision to stop answering it manually. Measure the call volume before and after; that number justifies the next step.

Digital Perpetual helps Singapore SMEs build enquiry handling and coverage systems that hold up through peak season. If your December plan currently depends on one person being reachable, that is the conversation to have in September.