Skip to content

Staff-to-task matching

The best person for a task isn’t always the one who happens to be free. Our model hands out the work by skills, preferences, cost and what each person should learn next, all at once.

  • Free 30-minute call
  • No commitment
  • You own the code
Typical industries
Healthcare, Field service management
Problem type
Resource planning
You receive
A documented model, run on your data

In short: A staff-to-task matching model assigns people to tasks or projects based on skills, preferences, development goals and cost together. It finds the best match for the whole team at once, rather than taking whoever is free first.

The problem

An urgent job lands on Tuesday morning. Your most senior engineer is free, so she gets it. The junior who needed exactly this kind of work waits another quarter. And the on-call weeks, when someone must be reachable outside working hours, land on the same three people again. Nobody chose that outcome. The spreadsheet just produces it, month after month.

Staff-to-task matching decides who does what. It looks easy until you write down everything that matters. Skills and certificates count. So do preferences, development, fairness and seniority, meaning how long someone has worked for you. A spreadsheet tracks only one thing well: who is free when. So availability ends up deciding everything.

Why availability-first matching fails

The usual method goes one task at a time. Find who’s free, drop the task on them, move on. If there’s time, you check skills. You hardly ever check who should learn what or whether the work is shared fairly. Whoever is free most often collects every unwanted job. Past a handful of people, there are far too many combinations to weigh by hand.

What changes

Today, you find who’s free, hand over the task and move on. Our model proposes the whole week’s assignments at once. Each task goes to someone qualified who fits it well. Harder tasks go to people who’ll learn from them. The unpopular jobs rotate. You review the plan, swap a name where needed and send it out.

What you get

Staff to task matching optimisation weighs skills, preferences, development, fairness and seniority together. Stretch work and unpopular jobs get shared out by rules everyone can see. We build the model on your skills matrix, the list of who can do what. We add your team’s preferences and the tasks that are coming up. Then we test it against your past assignments before it goes live. It has to work in the bad weeks as well, when a project is urgent and two people are off sick.

  • A model written around your own staff-to-task matching rules.
  • Results measured in the numbers you already track.
  • It slots into how you already work, and you can check every answer it gives.

Ranking candidates isn’t the same as matching

Most skill based assignment software ranks candidates for one task at a time. A typical planning add-on works like this. Skills make up most of the match score and spare capacity the rest. You pick a task, it suggests a handful of people, and you choose. It’s useful, and it still looks at one task at a time.

Take a small example, with scores out of ten. Anna scores 9 on task A and 8 on task B. Ben scores 8 on A but only 2 on B. Rank task A first and Anna wins it, so B goes to Ben: 11 points in total. Give A to Ben and B to Anna, and you get 16.

Each single ranking was correct. The total was poor because of the order the tasks came up in. A matching model scores the whole week at once and finds the 16. With two people you’d spot it yourself. With thirty people and eighty tasks, you won’t.

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.

Skill-based assignment

In optimisation, staff-to-task matching is called an assignment problem: pair people with tasks in the best way. The first criterion is skill. Every task has a requirement, such as a certificate or a security clearance. Other tasks need a pay grade or a level of experience. And among the people who meet the requirement, some do the task better than others. Skill-based assignment gives each task to someone who meets the requirement and does it well.

In the operating theatre, there is no choice. An operation needs a surgeon, an anaesthetist and a theatre team with the right specialty and qualifications. In field service the same applies, with many more jobs. Every job needs a technician with the right training who knows the spare parts. Otherwise a technician drives out to a job they can’t finish.

Preferences people can live with

Preference is the second dimension. People take on work they want and resist work they don’t. So a plan that ignores preferences stops working within the week. The model takes in wishes about location, shift pattern, project type and colleagues. It meets each wish as long as it doesn’t conflict with a hard constraint, a rule the plan must always keep.

Development and fairness

Development is the dimension most operations leave out. Stretch work is a task just beyond what someone is used to. It might be a complex case, a lead role or a new client. A senior who happens to be free rarely learns anything from it. If you hand it out yourself, you soon favour the same people. The same people get the good work, and the same juniors get the dull work.

A model makes development an explicit goal. It spreads stretch work across the team. The model counts in that such work may then be finished later. And it sets an upper limit per person, so nobody gets more than they can handle. Unpopular work such as on-call weeks, travel or the difficult customer is shared out the same way. When a team member asks “why me?”, you can show them the answer.

Seniority rules come on top. Some tasks need a minimum level of experience under your own policy, say five years in the job. Some assignments go only, or first, to the people who have been with you longest. The model enforces these as fixed rules or weighs them as goals, whichever your policy says. You don’t have to decide them on the spot while a task still has no name on it.

One model for the whole match

We solve the whole matching as one optimisation, with every dimension as a rule or a goal. We use the same method for workforce shift scheduling and for the general assignment problem. The method is called constraint programming. The model first rules out every assignment that violates a hard constraint. Only then does it search the rest for the one that best fits skill, preference, development, fairness and seniority.

Project staffing and team assembly

Project staffing optimisation picks a team from the people you have. The team must cover the skills the project needs and stay within budget. Consulting firms make these decisions every day. They match consultants to client projects by ability, development and utilisation. Utilisation is the share of a consultant’s hours billed to clients, say 80%. These decisions shape the result for the client and careers inside the firm. Team assembly optimisation makes them visible, and you can give reasons for each one.

Common questions.

Something else? Ask us directly.

Because picking the best person one task at a time doesn’t add up to the best week. Typical skill-matching tools rank a few candidates per task, mostly on skills and partly on spare capacity. That helps for a single urgent job. For a whole week of work, a matching model scores all tasks together and avoids using up your strongest people early.

A skills matrix is a table that shows who can do what. Start with what you can check: certificates, licences and clearances, with expiry dates. Then rate proficiency on a short scale, say 0 to 4, and agree what each step means before anyone rates anyone. Update it on a fixed rhythm, often every quarter. A matrix that’s three years old matches people to skills they used to have.

They’re more likely to when they can see the rules. Publish what the model looks at: skills first, then people’s wishes, what each person should learn next and a fair share of the unpopular work. You can overrule a proposal and note why. If a team member doubts that the plan is fair, you can show them how many on-call weeks each person has had. After a gut decision, you have no such number to show.

A stretch task is one where someone learns something new, so it takes longer. More stretch tasks mean more risk to the deadline, so set limits. Give each stretch task a risk rating. Then set how many stretch tasks one person or one project may have at most. Where the risk is high, an experienced colleague works on the same task. The model then hands out stretch tasks wherever your limits allow. You set the caps, and the model keeps to them every week.

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.

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