The visible cost: the wage-bill maths, step by step
Start with the hourly rate. £16.30 is not a payslip number — it is what an operator actually costs to employ in the UK in 2026: the National Living Wage, plus employer National Insurance at 15%, plus pension contributions at 3%, plus holiday cover. No supervision, no recruitment, no agency margin — just the baseline cost of one person, on shift, sorting parcels.
Now the hours. Take a typical parcel operation: two 10-hour shifts a day, 7 days a week, 50 weeks a year. Each staffed position on that pattern consumes:
- 20 hours a day (2 shifts × 10 hours), × 7 days = 140 hours a week
- 140 hours × 50 weeks = 7,000 staffed hours a year per position
- 7,000 hours × £16.30 = £114,100 a year per position
A manual sort of any real size needs a wall of those positions. At 12 sort positions — a realistic crew for feeding, sorting and clearing a mid-size destination sort — the annual bill is 12 × £114,100 = £1,369,200 a year. Not capital. Not one-off. Every year, rising with every National Living Wage settlement, for as long as the parcels keep coming.
"Position" means a role kept staffed across both shifts — so 12 positions is roughly 24 heads on the rota before absence cover. Rerun the sum with your own shifts and headcount; the arithmetic is the point, not the exact crew.
Why nobody sees this number
Almost no operation carries a budget line called "manual sortation". The cost hides inside a warehouse wage bill that also covers picking, packing, loading and goods-in, so the sort's share is never isolated. Agency spend sits with a different budget holder, overtime is approved shift by shift, and rework is booked — if it is booked at all — as a customer service cost. The first step in any automation case is simply consolidation: pull the sort's true hours out of the rota and price them at the all-in rate. In most operations that exercise alone changes the conversation, because a cost nobody owns is a cost nobody challenges.
The invisible costs on top
The wage bill is the number payroll can print. In most operations, the real cost of a manual sort runs meaningfully higher, through costs that never appear on the same page:
- Agency premiums at peak. Peak is precisely when sortation demand doubles and precisely when agency labour is scarcest. Agency rates typically carry a substantial margin over the £16.30 baseline — and peak agency staff sort slower and mis-sort more, because they are new.
- Overtime spikes. When a shift runs over to clear the belt, every one of those hours comes at premium rates. Overtime is the cost of having no headroom.
- Mis-sort rework. Manual sortation error rates look small as percentages, but at thousands of parcels an hour each error becomes a missed departure, a re-handle, a carrier claim or a customer refund. Rework is labour spent doing the same job twice, plus the commercial damage of the late parcel.
- Training and churn. Repetitive manual sortation is high-turnover work in most operations. Every leaver means recruitment cost, induction time and weeks of below-rate performance from the replacement.
- Absence cover. A missed shift on a manual sort is capacity that simply vanishes — the parcels do not wait. Cover comes from overtime, agency or a delayed departure, all of which cost money.
- Supervision. Twenty-plus operators need team leaders, quality checks and rota management — headcount that exists only because the sort is manual.
This is the pattern behind the problems operations directors actually report: missed shifts, agency workers, overtime spikes, dispatch bottlenecks, mis-sorts, rework and operator dependency at peak. None of them is a people problem. All of them are properties of a process that scales only by adding hands.
What the same money buys in automation
Here is the worked example, with the assumptions stated. An operation sorting 9,000 parcels per hour into 16 destinations with 12 manual sorters replaces the sort with an automated sortation system. The system costs £320.9k to buy outright — or around £6.5k per month on a 60-month finance term at 8% APR. It removes 10 of the 12 operators from the sort (two remain around the machine for induction and exceptions).
- 10 positions × 7,000 hours × £16.30 = £1,141,000 a year in labour removed
- Allow £12k a year maintenance on the system
- Bought outright: £320.9k against ~£1.1m a year of saving = payback in roughly 4 months
- Financed: £6.5k a month against roughly £95k a month of labour — cash-positive from month one
And the monthly payment buys more than removed wages. A sorter runs at the same rate in hour ten as in hour one, so the dispatch bottleneck at end of shift disappears. It reads a barcode instead of a label glanced at under pressure, so mis-sorts — and the rework chain behind them — fall away. Peak stops being an agency recruitment problem, because the machine's capacity is the same in December as in May; the two retained operators flex, the sort rate does not. Maintenance is a scheduled line item rather than an emergency, and the whole system is built in the UK to ISO 9001:2015 with UKCA/CE marking, with conveyor packages typically on site in 6–10 weeks. None of that appears in a payback calculation, and all of it shows up in the operation.
That last comparison is the one worth sitting with: £6.5k a month versus £95k a month for the same sort. The full derivation, including what drives system prices up and down, is in our guide to what warehouse automation costs in the UK, and choosing the right machine for the job is covered in the parcel sorter comparison. The ROI calculator reruns the whole model with your volumes, shifts and rates.
You are already paying for the automation — you're just not getting it
This is the argument that reframes the decision. An operation spending £1.37m a year on manual sortation is already paying an automation-sized bill — several times over — and receiving a process that mis-sorts, misses shifts and buckles at peak in return. The question is not "can we afford to automate?" but "what are we currently buying with this money?" With every Orion system available monthly — entry systems from around £4,000 per month, 60-month terms, subject to specification and approval — automation stops being a capital event and becomes a line item that is, in the worked example, roughly a fifteenth of the labour line it replaces. That is why finance changes the conversation: the money is already leaving the building every month. The choice is what it buys. The same logic applies whether you run a parcel centre, a 3PL or a fulfilment operation.
When manual sortation still wins
Honesty matters more than a sale, so: manual sortation is the right answer more often than automation vendors admit. It wins when volumes are genuinely low — if the day's sort is an hour's work for three people, a sorter would sit idle and the arithmetic above collapses. It wins when the operation is short-term: a pop-up site, a lease with two years left, a contract that may not renew — a 60-month finance term needs a 60-month operation. And it wins when variability is extreme: if every item needs human judgement — ugly freight, no barcodes, constantly changing rules — a machine adds cost without removing labour. There is also a middle path worth naming: many operations automate the predictable core of the sort — the 80% of parcels that are clean, barcoded and conveyable — and keep a small manual cell for the exceptions. That hybrid keeps the headcount low without pretending the ugly freight does not exist. The test is not ideology; it is the same arithmetic run honestly. If your labour line is small, flexible and error-tolerant, keep it. If it is seven figures and growing, the maths above is waiting.