Workforce operations

The Skill a Roster Can't See: Matching Support Workers to the People They Support

Published 9 October 2026 · about a 9-minute read

A worker can hold every credential a shift requires and still be the wrong person for that participant, on that day. Fit, how someone's pace, communication style and way of handling a hard moment sit against the particular person they are supporting, is a real dimension of skill. It rarely shows up on a roster, it almost never shows up in a training record, and when it is handled well it is usually because one coordinator has learned it over years and carries it in their head. None of what follows is about a system deciding who suits whom. It is about what a provider can actually do, with or without software, so that knowledge survives the person currently holding it.

Taieri Walsh, founder of CORA Workforce

Taieri WalshFounder of CORA.

Two support workers, on paper, are identical. Same induction, same manual handling sign-off, same first aid certificate, same completed training for the person's support needs. Put one of them in the house and the shift runs well. Put the other in the same house, with the same plan, and it does not, not because anything on their record is wrong, but because the two people in the room do not sit well together. One is quick, talkative, constantly initiating. The person they are supporting needs quiet and time to process. Nothing in either worker's file would have told you that in advance.

This is not a story about a bad worker. Both of them might be excellent, for the right person. It is a story about a dimension of skill that almost never gets written down anywhere, and that most providers only manage case by case, through whichever coordinator happens to carry the knowledge in their head that week.

Fit is a skill, not a soft extra

In short: a worker can hold every credential a support requires and still be the wrong person for that participant, that household, that day. Relational fit changes what the person actually receives on a shift, same credential, same training, a materially different outcome, which makes it a capability question rather than a personality quirk.

Someone who is genuinely excellent in a busy day program, quick, energetic, constantly engaging, can be exactly the wrong person for someone who needs quiet, predictability and low stimulation. That is not a personality clash in the way the phrase gets used about two people who do not get along in a staffroom. It changes what the participant actually receives on that shift. Same credential, same completed training, a materially different outcome, because fit is doing part of the work the support is meant to do. Treating it as an afterthought is exactly how a fully compliant roster still produces a bad shift for the person on the receiving end of it.

What "fit" actually consists of

In short: fit is not one thing, and it is rarely about whether two people simply get along. It is a specific set of dimensions, each of which can be observed, even though almost none of them are routinely written down anywhere a second person could read.

The dimensions that actually matter, in the situations providers see most often:

  • Pace and energy. A worker who moves and talks quickly can be exactly right for one person and overwhelming for another, independent of how skilled either worker is.
  • Communication style. Direct versus gentle, verbal versus visual, how much is explained versus simply done. A mismatch here can look like non-compliance from the person being supported when it is actually a style problem.
  • Tolerance for unpredictability. Some workers are calm when a plan changes mid-shift. Some participants need a worker who is calm about that far more than they need one who is highly qualified.
  • How someone handles escalation or distress. Staying steady, not taking it personally, not needing the moment to resolve quickly for their own comfort, is a specific and learnable skill that some workers have further developed than others.
  • Continuity and familiarity over time. Whether this is the fifth time this worker has supported this person, or the first, changes what a shift actually looks like before anyone says a word.
  • Cultural and language considerations. Shared language, cultural background or lived understanding of a community can be central to whether a person feels genuinely supported, not incidental to it.
  • Gender, where a participant has a preference. A stated preference about the gender of the person supporting them, for reasons of comfort, culture or history, is a choice to be respected, not a scheduling inconvenience.

None of these show up in a credential. All of them show up in the quality of a shift.

Why continuity matters

In short: every new worker has to be told, in some form, what the last one already knew. For a participant that is not a minor inconvenience, it is a repeated cost of re-explaining who they are, and high turnover makes a good match harder to hold onto even once it has been found.

A participant who sees a new worker every few weeks is not just losing a familiar face. They are repeatedly doing the work of explaining their communication style, their history, what helps and what does not, often at exactly the moment they most need support rather than a briefing. That cost sits with the person receiving support, not with the roster, and it is easy for a provider to underweight precisely because it never shows up as a line item anywhere.

Continuity is also what makes a good match worth anything. A worker and a participant who fit well together on day one are worth considerably more on day fifty, once trust, shorthand and shared history have built up. Rostering decisions that treat every shift as interchangeable, filling the slot with whoever is available this time, quietly erode that continuity even when nobody intends to.

Why it gets left off the roster

In short: fit gets left off the roster not because anyone decided it does not matter, but because it has no field to go in. A credential has an expiry date and a tick box. Fit does not, so it falls out of the system the same way anything without a field falls out, quietly and completely.

A credential is binary, exportable, auditable, the kind of thing a system can hold without an argument. Fit has none of that. It is not something you tick, and it rarely survives being reduced to a number without losing the thing that made it useful in the first place. So it gets handled informally, or not at all, and a provider that would never let a first aid certificate lapse unnoticed will happily roster on fit purely by memory, because there has never been anywhere else for that memory to live.

When it is done well, it rests on one person

In short: where fit is managed well, it is usually because one coordinator has built up years of knowing both the team and the people they support, and carries the whole picture in their head. That is genuinely skilled work. It also does not survive that person leaving.

A coordinator who has worked alongside a team for long enough to know how each worker actually operates, not just what they are qualified for, and who also knows the people they support well enough to picture how a particular pairing will go, can make a genuinely good call. That is skilled work, and it deserves to be named as such rather than dismissed as gut feel. But it does not scale past what one person can hold in their head, it is invisible to everyone else in the organisation, and the day that coordinator takes leave, moves teams or resigns, the knowledge leaves with them. Nobody notices immediately. The matches just get quietly worse, and nobody can point to why.

The structural problem for larger providers

In short: past a certain size, no one person can hold the whole picture. A new coordinator inherits a roster and a rough idea, and has to rebuild the rest slowly, by trial and error, and the people being supported absorb the cost of that learning curve while it happens.

Once the workforce and the number of people supported both grow past what a single memory can carry, the informal version of this stops working, not because anyone got worse at their job, but because the job itself changed shape. A new coordinator inherits a roster and a rough idea, and rebuilds the rest slowly, by trial and error, over months. The people being supported absorb the cost of that learning curve, because they are the ones living through the mismatched shifts while someone new works out again what the last person already knew.

What a provider can do, without any software

In short: none of this needs a system to start improving. Writing down what a person actually needs beyond their credential requirements, keeping that somewhere more durable than one person's memory, asking the participant directly, and recording what actually happens after a shift will do most of the work.

  1. Name what each person needs beyond the credential a shift requires. Not "needs a support worker," but the specific things that matter for them: pace, communication style, how they prefer distress to be handled, any cultural, language or gender preference. If this does not exist in writing yet, it is the real starting point.
  2. Write it down somewhere a second person can read, not only in one coordinator's head. It does not need to be sophisticated. A shared document that survives a staff change is worth more than a perfect system that depends on one person remembering to update it.
  3. Ask the participant. Directly, and plainly: what do you want in a worker, and what has not worked in the past. People are often not asked this at all, and the answer is usually clearer than anyone assumed.
  4. Record what is actually observed after a shift, rather than relying on recall weeks later. A short note at the time, this pace worked, this did not, is worth more than anyone trying to remember it when the next rostering decision comes up.

What CORA records, and what stays a human judgement

In short: CORA does not match, score or suggest a pairing between a worker and a participant, and there is no such feature anywhere in the product. What it holds is practice observations and supervision notes against a named worker, with the setting captured, and a plain-English paragraph built from that record. The decision about who suits whom stays a human judgement, in full.

Worth being direct about this before describing anything else, because the subject invites exactly the wrong kind of claim. CORA does not do personality matching. There is no algorithm, no score, no suggested pairing, and no ranking of which worker is the "best fit" for a given person. Anything that follows should be read with that firmly in mind.

What CORA does is more modest, and aimed squarely at the problem described above: knowledge that only ever lived in one coordinator's head. It records practice observations against a named worker, with the real setting the support happened in captured alongside them, a participant's home, a day program, a community outing, wherever the shift actually was, rather than a classroom or a module. It holds supervision notes against that same named worker, so there is a record of how someone works, not only that a supervision session took place on a given date. From that record, it builds a plain-English paragraph about the individual worker, something a team leader can actually read, or say out loud in a supervision conversation, rather than a table of percentages that tells nobody anything about the person.

None of that decides who goes on a shift. What it does is give the next coordinator, the one who has not yet built up years of knowing this particular team, something written to read before they roster anyone, instead of starting again from nothing the week the person who used to hold it all in their head leaves. That is the whole claim. The record, not the outcome.

See how CORA records the observable half of this

If what you know about fit currently lives in one coordinator's head, the fastest way to see the difference is to look at how practice observations and supervision notes get recorded against a named worker.

See how it works Book a demo

Frequently asked questions about matching support workers to participants

What does it mean to match a support worker to a participant?

It means choosing a worker for a particular person based on how the two actually work together, not only on whether the worker holds the right credentials and is free for the shift. Pace, communication style, how someone handles distress, and continuity over time all affect the quality of a shift between two people who are each individually qualified. Fit is a dimension of skill, not a personal preference on either side.

Is matching support workers to participants required under the NDIS Practice Standards?

The Practice Standards do not name a matching or compatibility process. What they require is that supports are person-centred and that a participant's choices, including about who supports them, are respected and responded to. In practice that means a provider needs some way of knowing and acting on what a particular person needs beyond a worker's qualifications. Providers should confirm the detail that applies to their registration groups with the NDIS Quality and Safeguards Commission.

Does CORA match support workers to participants automatically?

No. CORA does not score, rank or suggest a pairing between a worker and a participant, and there is no matching or compatibility feature of any kind. What it holds is practice observations and supervision notes recorded against a named worker, with the setting captured alongside them, and a plain-English paragraph about that worker built from the record. That gives a coordinator something written to read before they decide. The decision about who suits whom stays a human judgement.

Why does continuity of support worker matter for a participant?

Because a new worker has to be told, every time, what the last one already knew. For a participant that can mean repeating personal history, explaining communication needs, or re-establishing trust before the support itself even starts. High turnover does not just cost a provider in recruitment, it costs the participant in the work of re-explaining themselves, and it makes genuine fit harder to hold onto even when a good match has already been found.

What can a provider do about fit without buying any software?

Four things, none of which need a system. Write down, for each person supported, what they actually need beyond the credential a shift requires. Keep that written somewhere a second person can read, not only in one coordinator's memory. Ask the participant directly what they want in a worker, rather than inferring it. And record what is actually observed after a shift, rather than relying on anyone's recall weeks later when a rostering decision comes up again.

This guide is general information for NDIS providers, not legal or compliance advice. Always check the current requirements directly with the NDIS Quality and Safeguards Commission, because the detail does change.

Sources and further reading

← Back to all guides