Adding a second plant rarely doubles capacity in practice. It adds a coordination problem, and how that
problem gets solved determines whether the second site is an asset or an expensive complication.

The Three Coordination Questions

Which site makes what. Sometimes determined by capability — only one plant has the equipment or
certification. Often it’s genuinely open, and that’s where planning has leverage.
How load gets balanced. When one site is overloaded and another has capacity, moving work is obvious in
principle and often blocked in practice by tooling, qualification, or local reluctance.
How shared resources get allocated. Common materials, shared tooling, specialist labour — when these are
scarce, someone has to decide who gets them, and if that decision isn’t made deliberately it gets made by
whoever asks first.

Centralised, Local, or Between

Fully centralised planning optimises the network but tends to produce schedules that ignore local realities,
which then get overridden until the central plan is decorative.
Fully local planning respects site knowledge and produces network-level waste — one site running
overtime while another runs short.
The workable structure is usually central allocation and local sequencing. The network level decides which
site produces what volume in which period, based on capacity, cost, and logistics. Each site sequences its
own work, because that requires knowledge no central system holds. The boundary needs to be explicit —
most multi-plant friction comes from ambiguity about who decides what.

Comparing Sites Honestly

Allocation decisions need comparable data, and this is where most multi-plant planning breaks down first.
Demonstrated capacity by site, measured the same way at each. Cost per unit including labour, overhead,
and yield differences. Freight cost to serve from each location. Lead time to the customer. Qualification
status — which sites are approved for which products and customers, which is frequently the binding
constraint and frequently out of date.
If sites define capacity or cost differently, comparison is impossible and allocation defaults to habit.
The organisational obstacle
Plant managers measured on their own site’s efficiency and cost will resist taking on difficult low-volume
work and will hold onto profitable runs. Both behaviours are rational given the measurement and both
damage the network.
If you want network optimisation, measure at network level and adjust site metrics for allocation decisions
made centrally. A plant given the hard work shouldn’t look worse for it. This is a measurement design
problem and it’s usually the real barrier — not the planning system.

Where to Start

Establish comparable capacity and cost data across sites before attempting to optimise anything. Then
maintain an accurate qualification matrix — which products can genuinely be made where. Most
manufacturers find this document is significantly out of date, and it constrains every allocation decision
until it’s fixed.
Qualifying additional site-product combinations is also one of the cheapest resilience investments available.
Each one added widens your options during a disruption and during ordinary load imbalance, and the cost
is engineering and validation time rather than capital.
ticktick.ai plans allocation across sites using comparable capacity and cost data with a live qualification
matrix, leaving sequencing to each plant.