Workforce allocation across sites
Each site plans its own rota, and agency staff cover the shifts it can’t fill itself. Our model places your staff across every site and shift at once.
- Free 30-minute call
- No commitment
- You own the code
In short: A multi-site workforce optimisation plans staff across all your locations and shifts in one model. It sends people to the sites that need them, as long as the commute isn’t too long. It covers absences from a shared pool, so fewer shifts go to agency staff.
The problem
At twenty to seven, a nurse at Site A calls in sick. You ring the other sites one by one. Nobody calls back in time, so you book an agency nurse. Across town, a nurse with the same training has little to do on a ward with few patients today.
With one site, the rota says when each person works. From the second site on, you also decide where each person works. With hundreds of staff across many sites, that choice decides travel cost, how many people work each shift and how often the same people get sent elsewhere.
Signs your sites are planned in silos
Site managers call each other every morning to borrow staff. You spend more and more on agency staff, though on paper you employ enough people. The same people keep getting sent across the region while closer colleagues stay put. And nowhere can you see at a glance which staff with the right training are free today.
Why site-by-site planning fails
The usual setup is one spreadsheet per site and a phone call when a site runs short. Each manager fills their own rota, and the shifts they can’t fill go to the agency. No site manager sees that the next site has exactly the cover that’s missing. So one site pays agency rates for a shift while another has staff with the same training and nothing to do.
What changes
Today, each site fills its own rota, and you phone around when a shift falls short. Our model places people across all sites and shifts at once. When a nurse calls in sick, it re-plans with all available staff and proposes the nearest replacement with the right training. You confirm the swap instead of ringing round all morning. Every placement fits the person’s qualifications and keeps the site’s rules.
What you get
Multi-site workforce optimisation plans all your sites together. It solves one assignment problem: which employee works which shift at which site. Staff who are free at one site and have the right training become visible to every other site. You stop paying agency rates for shifts your own people could have covered. You see the agency cost reduction item by item: which agency shifts you no longer book. We build the model on your demand forecasts, where people live, travel times, qualifications and absence rules. We test it before a single shift is rostered.
- A model written around your own workforce allocation across sites rules.
- Results measured in the numbers you already track.
- It slots into how you already work, and you can check every answer it gives.
Can she work at the other site tomorrow?
Rest times matter most when people move between sites. A person who works a late shift at one site can’t always take an early shift at another the next morning. Your rest times between shifts go into the model as your own rules, with any exceptions your contracts allow. Every move it proposes keeps them. Today, a too-short rest often shows up only once staff already have their rota. With the model, you see it before you confirm the swap.
Take a nurse whose late shift at Site A ends at 23:00. The early shift at Site B starts at 07:00. That’s eight hours of rest. If your rules allow that only once a week, she can do it once this week and not twice. On the phone at twenty to seven, you rarely have her whole week in view. The model checks it on every swap it proposes.
How it works
We write your resources, the work they have to cover and your rules into one model. You get a plan that says which resource does what and when. It’s the plan that fits your goals best, such as the lowest cost or the shortest waiting time. Every rule is kept at the same time, and that’s the part a spreadsheet can’t do.
Cross-site demand and travel cost
The core is the assignment problem: pair people with shifts so that the total cost is as low as possible. Here it covers all your sites, not just one. Each employee has a home location, qualifications, travel costs and availability. For each site, your cross-site demand forecast says how many people of each role every shift needs. Each site also has its own rules on required training.
Travel cost is part of the total cost the model keeps as low as possible. A guard who lives near Site B covers the late shift there. Then nobody has to cross the city from Site A. Site-specific qualifications restrict the matching: a nurse needs the certificate the ward requires, a guard the clearance and a driver the licence. The model only places people where they’re allowed to work.
Floating pools and absence rebalancing
A floating staff pool is a group of people who belong to no single site. They help out wherever work piles up or staff are missing. That makes the pool your buffer, a reserve for absences. Without it, every site keeps its own spare staff for its worst day of absences. One mobile team can replace much of that spare staff. When a person calls in sick, absence rebalancing re-plans with all available staff. It pulls a floater with the right training instead of a rushed agency booking. Or it takes a colleague from a nearby site that has one more person on shift than it needs.
Multi-period allocation plans many shifts in a row, not each one alone. That’s where the rest rules above come in. Who you place this morning changes who’s rested and free tonight. Cross-site scheduling optimisation decides all of it in one model.
One model for the whole network
Demand, travel, qualifications, rest and absence become constraints and objectives in one model. A constraint is a rule the plan must keep, such as a required licence or a minimum rest between two shifts. The method is called constraint programming. The model first rules out every placement that would violate a rule. Only then does it search the rest for the cheapest plan across all sites. When a person calls in sick or a site needs more staff, it re-plans. It’s the same machinery as employee rostering, except the model picks the site as well as the shift.
Where it matters most
Healthcare is the acute case. Wards across hospitals and care homes need a set number of nurses on every shift, for example four at night, at least one with intensive-care training. Sending a nurse from a ward with spare staff costs far less than bank or agency staff. Bank staff are the hospital’s own pool of temporary cover. Multi-location staff rostering turns that into routine.
Security and facility services work with the same maths across the buildings of a city. Guards and cleaners are cleared for specific sites, and demand swings through the day. Whether a contract makes a profit depends on one choice. Do you send the cleared colleague nearby, or pay for agency cover?
Common questions.
Something else? Ask us directly.
Big enough to cover the absences you can predict, and small enough that you don’t pay floaters to wait for work. Floaters are staff without a fixed site who step in wherever someone is missing. We start from your own data: how many shifts per week did you fill with agency staff or overtime last year, by role and site? The model then tests pool sizes against those weeks and shows the cost of each. It sizes the pool per role, because absence rates differ from role to role.
That’s a contract question first: what does each contract say about the place of work? Some allow work at other sites within a region, and some don’t. We put whatever the contracts say into the model as a rule per person, so the plan only proposes moves people have agreed to.
It lists everyone qualified for the shift, then rules out anyone who would get too little rest or go over their maximum hours. Among the rest, it weighs travel time, cost and how often each person has been moved lately. You get a short list with the best option on top and confirm one.
It can, if the same people always go. That’s why the model counts the moves per person and spreads them evenly. It also meets people’s wishes wherever the rules allow. A floating pool of people who chose to float takes on most of the moves, so everyone else moves less often. When everyone can see the move counts per person, the conversation gets easier.
Sometimes it can, and if that’s the case, we’ll tell you. Standard software is built for an average operation, though. Yours runs on its own rules: your contracts, your exceptions, the things you know by heart. A custom model writes those rules in, and that’s where the better plan comes from.
You don’t need either to start. We build the model from the data you’ve already got, even if it lives in three spreadsheets and one person’s head. You get a system you run yourself. We train your team until they no longer need us.
You see a first result with the prototype, before you commit to anything bigger. We start with a free 30-minute call about your operation and the model that would fit it. Then we build a working prototype and run it on your numbers.
Similar pages.
Terms on this page
Every abbreviation explained in plain words, in our article “Metrics and methods”.
Tell us what you’re planning.
In a free 30-minute call with a specialist, you find out whether optimisation can help.
- Free, no commitment
- You talk to a specialist directly
- A clear go or no-go