Eleven Australian Commonwealth agencies have roughly A$485 million in published contracts naming VMware as supplier, expiring across 2027 and 2028. All were signed after Broadcom completed its acquisition, and 79% by value was signed in 2025 or later.
The number matters less than what happens next. Each of those agencies will face the same decision: renew, move some workloads, or move the lot. I wanted to know what the public record could tell a buyer approaching that choice — and, just as usefully, what it couldn’t.
Three things came out of it. Contract consolidation makes published values badly misleading if you compare them the obvious way. The expiries cluster hard in 2028. And the Australian hiring market shows far more demand for VMware skills than for any alternative — which says something about how practical a migration threat is as a negotiating position, though less than you might first think.
When the renewals land
Nearly four-fifths of the value sits in 2028. For a buyer that means a lot of organisations reaching the same decision in the same window — which affects how much attention any one of them can expect to command, and how much competition there is for whoever would do the work.
What the money actually says
The obvious way to read these contracts is the wrong way, and Defence shows why.
Twenty-seven Defence VMware contracts ended on 11 December 2025, with a combined published value of A$95.9 million. One contract worth A$178 million began the next day, running three years.
Set against Defence’s largest earlier contract, the new agreement looks like a twelvefold increase. Set against all 47 contracts it replaced, annualised contracted value is slightly lower. The consolidation is the story; the headline number is an artefact of how procurement was structured.
If you are benchmarking your own position, compare total annualised contracted value across all agreements, not contract against contract. The published values invite the wrong comparison.
Across the four agencies where earlier contracts are visible, the same normalisation gives this:
That is a four-agency matched cohort, not a Commonwealth-wide figure. Those four hold 70% of current value, but the other seven are simply absent from the comparison — so this says nothing about Commonwealth VMware cost inflation generally, and I would not use it that way.
Is leaving a credible alternative?
Renewing under high switching costs is not evidence of satisfaction. Migrating a production estate is difficult, carries real risk, and competes for people who are already committed elsewhere. Renewal in those circumstances is frequently the rational decision.
It is partly testable, because migrations need people before they need contracts.
Two things stand out. VMware exceeds Nutanix seventeen to one. And Microsoft, not Nutanix, is the larger alternative in these advertisements — Hyper-V appears roughly eight times as often as Nutanix, which runs against most migration commentary I had read. The VMware roles being advertised are predominantly VMware Cloud Foundation administrators: positions supporting continued operation rather than an exit.
Measured against two other markets, Australia’s alternative-platform share is the lowest of the three:
This is weaker evidence than it looks, and worth handling carefully. Recruitment is an indirect signal. A migration may be delivered by a systems integrator, an existing partner, an offshore team or redeployed internal staff, none of which produce an advertisement — and vacancies are often written as “infrastructure modernisation” without naming a destination. The three markets also use three different job boards with different employer populations, so a similar ratio does not establish a similar market.
The defensible reading is narrower than “nobody is migrating”: Australian advertisements show substantially more demand for VMware skills than for the alternatives, which is consistent with continued dependence. Whether migration programmes are under way, this cannot say.
Three ways the decision goes
A renewal of this kind usually resolves one of three ways. The evidence does not tell you which to choose — that depends on your estate — but it does bear on the constraints.
Renew and optimise. Stay, and attack the bill through rationalisation: consolidate agreements, reclaim unused entitlement, right-size, negotiate term and uplift caps. Defence is the clearest public example of consolidation changing a commercial position.
Selective migration. Move what moves cheaply, keep what doesn’t, reduce dependency over time. This is where the skills picture bears most directly — a smaller programme can be staffed from a thin market where a full transition cannot.
Full platform transition. Replace across the estate. The public evidence is least encouraging here, not because the alternatives are weak but because the delivery capacity to execute at scale is hard to see in the sampled recruitment data, and the 2028 cluster means many organisations may be competing for the same scarce people at once.
What the evidence genuinely supports is narrower than a recommendation. You now know what eleven large Australian buyers contracted and when each comes up — a floor of public fact to negotiate from, reproducible from the dataset below. And if your leverage rests on a credible threat to migrate, that threat needs people behind it. An alternative you cannot staff is a weaker card than it looks across a table.
Forecast
A prediction that cannot be settled objectively is not worth making, so here is the resolution rule alongside the call.
Of the 19 contracts listed below, more than half by count will be followed, within 12 months of their expiry, by a new AusTender contract notice from the same agency naming VMware or its successor entity as supplier. Confidence: 65%. Resolves: 28 February 2030, from AusTender records.
The resolution date is set by arithmetic rather than preference. The last contract in the set expires on 21 December 2028; a twelve-month observation window runs to 21 December 2029; and agencies have 42 days to publish a contract notice. Anything earlier would leave seven of the nineteen contracts unresolvable.
This post first gave the resolution date as
Counting contracts rather than agencies or dollars is deliberate: it is least sensitive to one large renewal dominating the result.
Alongside it I will track Nutanix plus Proxmox as a share of advertised VMware roles in Australia, currently 11%. A durable rise above 15% would weaken the forecast — as evidence, not proof, since it is an indirect measure.
One thing remains unexplained. A$379 million concentrated in 2028 across a small number of agencies is either an artefact of contract terms or something more deliberate in how these agreements were structured. The data does not distinguish between them.
If you have been through a major platform renewal — how often was migration a genuine commercial alternative, and how often was renewal effectively the only practical choice? That is the question public records cannot answer.
Method
Contract data: the AusTender OCDS API, the Commonwealth’s contract notice register. Current contracts: end dates 1 January 2027 to 31 December 2028, UNSPSC families 4322/4323 and 81111/81112, value A$500,000 and above — 1,050 contracts totalling A$4.57 billion, of which the VMware-supplied subset is reported here. Earlier contracts: searching the approximately 110 days before each current contract’s start returned 92 VMware contracts, of which the 77 ending before 2026 (A$232 million, 10 buyers, 18 December 2023 to 14 December 2025) are treated as predecessors.
Annualised figures are contract value ÷ duration × 365.25, summed per agency. Where terms overlap, this slightly overstates contracted value in any single month.
Job counts: exact-phrase searches, 14–17 October 2026. Australia via Seek (VMware 239, Hyper-V 117, Nutanix 14, Proxmox 12), United Kingdom via Reed (212 / 158 / 16 / 24), United States via Dice (2,412 / 554 / 322 / 96). Each board has different coverage and employer mix, so absolute counts are not comparable across markets — only ratios within a market, and those only indicatively. A single measurement indicates position rather than direction; these figures will be re-measured monthly and published as a series.
Supplier matching is on the name recorded in the contract notice, which covers two entities: VMware Australia Pty Ltd (12 contracts) and VMware International Unlimited Company (7). One earlier record naming “VMware User Group Inc” — a training membership, not the vendor — was excluded.
The API returns a single release per contract notice reflecting its current state; amendment history is not exposed. A value shown here may therefore have been revised since the notice was first published, and no amendment lineage could be reconstructed.
What this evidence cannot establish
Published contract value is not expenditure. The Department of Finance is explicit that AusTender values are contracted amounts rather than money spent, and that they change through amendment. Every figure here is contracted value.
It is not a price comparison. Published values do not disclose cores, sockets, virtual machines or licence counts. No unit rate is derivable, and nothing here should be read as one agency negotiating better terms than another.
This is a supplier-identified subset, not all Commonwealth VMware procurement. It captures notices where VMware is the named supplier. Licences bought through a reseller, or bundled inside a larger systems-integration contract, do not appear — so the true total is higher than A$485 million by an unknown margin.
Some contracts bundle other products from the same vendor group. The Taxation Office’s largest is described as “CA Mainframe and VMware software licences”, the Reserve Bank’s as “Software licensing and support — CA Broadcom”. These cannot be read as pure virtualisation spending, and whether bundling is better or worse value than separate licensing is not determinable from published figures.
Scope is not constant. Estates grow and requirements change. One earlier Defence contract is described as “Computer Equipment”. A larger figure may reflect a larger purchase.
Seven of eleven agencies have no earlier contract visible, including the Taxation Office, the second-largest current buyer. Earlier purchases may have been made through a reseller, or fall outside the search window used. That is a gap in the data, not a finding about those agencies.
This is public-sector data only. Commonwealth agencies must publish contracts above A$10,000; private companies publish nothing comparable, so none of this can be extrapolated to them.
Contract detail
Every contract behind this analysis, by notice number, so any figure here can be checked against the source record.
| Notice | Agency | Value | Start | Expiry |
|---|---|---|---|---|
| CN4211096 | Department of Defence | A$178,086,240 | 2025-12-12 | 2028-12-11 |
| CN4142919 | Australian Taxation Office | A$108,625,000 | 2025-03-22 | 2028-03-21 |
| CN4071834 | Services Australia | A$93,973,803 | 2024-06-13 | 2027-06-22 |
| CN4175970 | Department of Defence | A$27,564,423 | 2025-06-30 | 2028-06-29 |
| CN4154952 | Reserve Bank of Australia | A$18,324,640 | 2025-04-29 | 2028-04-28 |
| CN4137626 | Department of Foreign Affairs and Trade | A$18,142,957 | 2025-02-13 | 2028-02-12 |
| CN4215000 | Bureau of Meteorology | A$12,193,449 | 2025-12-22 | 2028-12-21 |
| CN4150499 | Department of Employment and Workplace Relations | A$9,790,528 | 2025-05-12 | 2028-05-11 |
| CN4075434 | Services Australia | A$3,960,000 | 2024-06-23 | 2027-06-22 |
| CN4260252 | Australian Taxation Office | A$3,025,003 | 2026-06-30 | 2027-06-29 |
| CN4042775 | Department of Defence | A$2,964,192 | 2024-03-10 | 2027-03-09 |
| CN4150498 | Department of Employment and Workplace Relations | A$1,588,950 | 2025-05-12 | 2028-05-11 |
| CN4086032 | Department of Defence | A$1,505,556 | 2024-07-30 | 2027-07-29 |
| CN4180119 | Geoscience Australia | A$1,378,982 | 2025-08-18 | 2028-08-17 |
| CN4236283 | Department of Defence | A$1,097,704 | 2026-05-10 | 2028-12-11 |
| CN4175583 | Australian Transaction Reports and Analysis Centre | A$997,955 | 2025-07-29 | 2028-07-13 |
| CN4181064 | Federal Court of Australia | A$903,870 | 2025-07-16 | 2028-07-15 |
| CN4172594 | Australian Criminal Intelligence Commission | A$646,309 | 2025-07-26 | 2028-07-25 |
| CN4224787 | Services Australia | A$522,243 | 2026-01-30 | 2027-06-22 |
These are my personal views, based on the public evidence cited. They are not the views of my employer, and no part of this analysis uses non-public information.