Why placement management outgrows spreadsheets

Almost every placement operation starts in a spreadsheet, and almost every one outgrows it at roughly the same point. The symptoms are consistent enough to be diagnostic.

A spreadsheet is a perfectly reasonable tool for tracking twenty placements managed by one person. It stops being reasonable when three parties need the same information at the same time, and none of them can see it without asking someone.

The symptoms of outgrowing manual process

• Nobody can answer "where is this student up to?" without checking with a specific person
• The same information exists in three places and they disagree
• Preparing for audit means assembling records rather than exporting them
• Students find out about changes last
• Host facility capacity is known informally, in someone's head
• A staff member leaving takes institutional knowledge with them

If more than two of those are true, the constraint is the process rather than the people.

What placement software actually needs to do

Shared status, not status reports

The core function. Every party sees the same current state without anyone producing an update. The RTO sees their cohort. The facility sees students on site. The student sees their own progress. Nobody emails to ask.

Structured stages

Free-text status fields degrade immediately. Defined stages — introductory call, finding a workplace, applied, awaiting date, appointment booked, interview, awaiting start, started, completed — make progress measurable and bottlenecks visible. When six students sit at "awaiting date" for three weeks, that is a problem you can see rather than one you discover.

Matching that considers more than availability

Manual matching optimises for whoever is free. Good matching weighs qualification fit, host capacity, travel distance, availability and any RTO-specific conditions simultaneously. This is where AI-assisted scoring earns its place — not as a gimmick, but because a person cannot hold that many variables across hundreds of combinations.

Documents attached to people, not folders

Every document filed against the student it belongs to, with expiry tracking, and exportable as a set. This single change is usually what turns audit preparation from a fortnight into an afternoon.

Geography as a first-class input

Travel distance is one of the strongest predictors of whether a placement completes. If your system does not know how far a student is from a host, it cannot help you avoid the placements most likely to fail.

What it changes for each party

For students

Certainty. Knowing which stage you are at, when you start, where to go, how long it takes to get there and what documents are outstanding removes most of the anxiety of placement. It also removes the phone calls asking for that information.

For RTOs

Visibility and evidence. Live cohort status without chasing, and an audit trail that assembles itself. Compliance stops being an annual scramble and becomes a by-product of the process.

For host facilities

Less administration and more control. Set capacity per site, publish rosters once, notify automatically, keep documents in one place, and reach students directly rather than waiting.

Where AI genuinely helps — and where it does not

Worth being precise about, because the term is used loosely. The defensible application in placement is matching and prioritisation: scoring a large number of student–host combinations against multiple weighted criteria faster and more consistently than a person can.

What AI should not do is make the decision. A person confirms the match, the agreement is signed by people, and the supervision is done by people. The software narrows the field; it does not replace judgement.

Choosing a system

1. Can all three parties see the same state without asking anyone?
2. Can you export a complete per-student evidence file?
3. Are stages structured, or is status a free-text field?
4. Does it handle multi-site facilities with capacity per branch?
5. Do students get notifications, and is acknowledgement recorded?
6. Is there a mobile experience? Students and floor supervisors are not at desks.
7. Can you get your data out if you leave?

What to be sceptical about

The placement software market has a fair amount of noise in it. A few claims worth pressing on:

• "Fully automated matching." Matching can be scored automatically. Agreements, supervision and judgement cannot. Ask what a human still does.
• "Guaranteed compliance." No software makes an RTO compliant. It can make evidence easier to produce. The obligation does not transfer.
• "AI-powered" with no explanation. Ask specifically what the model does, what inputs it uses, and what happens when it is wrong.
• Impressive dashboards with no underlying process. A dashboard reflects data quality. If the data is entered inconsistently, the dashboard is confidently wrong.

Data you should be able to get out

Vendor lock-in is a real risk in a system that holds several years of compliance records. Before committing, establish:

• Can you export all student records, including attached documents, in a usable format?
• Is the export self-service, or does it require the vendor to run it?
• What is the notice period and what happens to your data afterwards?
• Where is the data hosted, and does that satisfy your obligations?
• What is the backup and recovery position?

Ask for an export during the trial rather than taking the answer on trust.

Adoption is the hard part, not selection

Most failed implementations fail on adoption. The system is bought, partially populated, and used inconsistently alongside the spreadsheet it was meant to replace — which produces the worst of both.

What tends to work:

1. Pick one cohort and run it entirely in the new system rather than partially across all
2. Turn the old process off for that cohort. Parallel running becomes permanent.
3. Decide who owns data quality, by name
4. Train the people who enter data, not just the people who read reports
5. Fix the friction points in the first fortnight, while goodwill lasts
6. Only expand once one cohort has run end to end

Mobile matters more than it appears

The people who most need to interact with a placement system are not at desks. Students are commuting or on shift. Floor supervisors are on the floor. Facility managers move between sites.

A system that assumes a desktop browser gets used by administrators and ignored by everyone else — at which point the data reflects only what administrators entered, and the shared visibility that justified the purchase does not materialise. Test the mobile experience with an actual student and an actual supervisor before deciding.

The honest caveat

Software does not fix a broken process, it accelerates it. If host relationships are weak, no portal creates capacity. If supervision is inadequate, no dashboard makes placements compliant.

What it does is remove the administrative drag that stops good operators from scaling — and make the parts that are working visible enough to build on.