Buying Corporate Training in Australia in 2026: A Procurement Playbook

The market size figure, and why it misleads buyers

Ken Research put Australia's corporate training market at USD 3.5 billion in 2023, on track for USD 5.4 billion by 2028. Good headline. Weak procurement input. What the figure hides is that a large slice of it is compliance training nobody chose, priced per seat on terms nobody negotiated, renewing on autopilot because switching feels worse than the invoice.

If you run L&D or procurement at an Australian enterprise, the number you need is your own cost per converted employee. Not seats. Not completions. Converted, meaning people who can now do work they couldn't do twelve months ago, at a standard somebody independently checked. Most Australian buyers can quote their platform licence to the dollar and go quiet on the second question.

Related reading: Australia's B2B EdTech Market in 2026: Who's Buying and Who's Winning · Australia's AI Talent Shortage in 2026: What Employers Actually Pay · University-Industry AI Partnerships in Australia in 2026 · the New Zealand version of this vendor playbook.

Public money is setting your floor price

The $12.6 billion National Skills Agreement, plus roughly $1.5 billion committed to keep about 500,000 fee-free TAFE places running through 2026, changes what any commercial vendor can charge for foundational content. If a competent version of your introductory data course is available free through TAFE NSW or TAFE Queensland, the paid version has to be better in a way you can name, or cheaper, or bundled with something a public provider can't offer. Assessment against your own internal role framework, for instance.

Vendors hate this question and buyers underuse it. Ask every provider quoting foundational training how their content differs from the fee-free equivalent. The strong ones have an answer ready and it almost always involves contextualisation to your industry and your systems. The rest talk about production values and mobile responsiveness.

There's a second-order effect. Because the public system carries so much of the entry-level load, Australian commercial vendors have drifted upmarket into applied and role-specific work, which is where their margin now sits. That's good for you if you're buying applied capability and bad for you if you assumed the market was still competing hard on price for basics.

Vendor tiers and what each one is good at

Five tiers, and buyers routinely shop in the wrong one. Rough 2026 price signals below, drawn from Australian enterprise deals at reasonable volume. Treat them as negotiation starting points rather than rate cards.

The common error is buying tier one and expecting tier five outcomes. A content library is a distribution mechanism. It is not a capability programme, and no amount of curation turns it into one. If your board asked for AI capability and you responded with a marketplace licence, you have bought yourself a completion dashboard.

The revised RTO Standards changed your due diligence

The Standards for Registered Training Organisations 2025 took effect on 1 July 2025, with sharper expectations on student outcomes, industry alignment and inclusive delivery. If you're buying accredited training, that hands you leverage you didn't have in 2023. ASQA has also flagged the marketing and recruitment of international learners as a critical risk priority for 2026, which tells you exactly where the regulator is looking.

The practical version is short. Ask every RTO on your shortlist for their most recent ASQA audit outcome and their completion rate broken down by cohort, not the blended average. Providers running a healthy business send both within a day. Providers who need a week to "pull that together" have told you something without meaning to.

One more that costs nothing: ask who owns the assessment materials at the end of the contract. A surprising number of Australian training agreements leave the buyer with no rights to the very framework they paid to have built, which turns your second-year renewal into a hostage negotiation.

What the APS mandate did to pricing

The APS AI Plan requires every Commonwealth public servant to complete foundational AI training, with agencies appointing a Chief AI Officer by July 2026. That's more than 200,000 people entering the training market at once, on government procurement timelines, in a labour force of roughly 14 million.

For private buyers the effect has been quietly irritating. Foundational AI literacy content is now abundant and cheap, because half the market built it chasing the public-sector wave. Applied, role-specific AI capability is scarcer and dearer than it was two years ago, because those same vendors moved their senior facilitators onto government delivery. If your 2026 plan assumes experienced facilitators at 2024 day rates, rebuild the plan.

Microsoft's A$25 billion Australian commitment, which includes helping three million Australians build workforce-ready AI skills by the end of 2028, pushes the same way. Abundant free supply at the bottom, constrained expert supply at the top. Buy accordingly. Take the free literacy layer without guilt, and spend real money only on the applied layer.

Here's the part I'd argue with most L&D teams about. The TalentLMS 2026 report found that only about 45% of L&D professionals have experimented with generative AI themselves, while nearly all of them are being asked to build AI capability for everyone else. You cannot specify a programme in a domain you've never touched. Before the RFP goes out, make your own team spend a fortnight actually using the tools. It's the cheapest quality control available and almost nobody does it.

Five questions that separate real vendors from good decks

• What percentage of your last three Australian enterprise cohorts finished, and in which week did drop-off spike?
• Show me a role capability framework you built for a client in my industry. Redact the name, keep the structure.
• Who facilitates, by name, and are those people available in my delivery window or are they currently on a government contract?
• What happens contractually if fewer than 70% of enrolled staff complete?
• Which of your competitors would you recommend if we were half our size?

That last one is unfair and it works. Vendors who answer it straight are usually the ones who don't need your logo badly enough to oversell you. I've watched a Melbourne insurer reorder an entire shortlist on the strength of one honest answer to it.

The renewal trap

Australian training contracts have a habit that costs buyers real money: the auto-renewing three-year platform agreement with a per-seat price that never falls as headcount grows. You sign at 400 seats, you grow to 900, and the unit price sits exactly where it was because nobody built a volume step into the schedule.

Two fixes, both easy to get if you ask before signature and nearly impossible afterwards. Put volume tiers in the pricing schedule with automatic step-downs at defined seat counts, so the discount triggers without a renegotiation. And cap the annual uplift, because CPI-linked escalation clauses in a market where content costs are falling is a one-way ratchet in the vendor's favour.

The other trap is the seat definition itself. Some Australian vendors count every person with an account, active or not, which means the 300 staff who logged in once in 2024 are still on your invoice. Define a seat as an active user in the billing period and you'll find the same contract costs 20% less without a single change to what you're actually buying.

A large Queensland services business I know found A$180,000 a year in exactly that clause during a routine renewal review. Nothing clever, no leverage, no threat to leave. They simply read the definition of a billable user for the first time in four years.

Where I'd put an A$400,000 budget

Assume a 1,200-person Australian company, mixed office and field workforce, a board asking pointed questions about AI. My allocation would look roughly like this. A$60,000 on a content library covering compliance and literacy for everyone, because that layer is commoditised and shouldn't consume management attention. A$240,000 on applied cohorts for the three role families where the work really changes, which in most Australian mid-market firms turns out to be engineering, finance analysis and customer operations. A$60,000 on internal capability, meaning two people whose actual job is programme design rather than vendor wrangling. The remaining A$40,000 stays unallocated until month seven.

That unallocated slice is the first thing finance cuts and the thing you'll miss most. Every training programme I've seen run at Australian scale needed a mid-year correction, and the ones with no budget for it simply carried the mistake through to December and called it a learning.

If you take a single thing into your next vendor meeting, make it the completion-and-drop-off question. Cheap to ask, hard to fake, and it reorders shortlists faster than any weighted RFP scoring matrix I've ever been shown.