VMware Migration Statistics 2026: Every Survey Normalised

86% ^S01 of surveyed VMware decision makers (CloudBolt-sponsored, n=302 ^S01, North America) are actively reducing their VMware footprint, with no time bound (fielded January 2026). In the same survey, 4% ^S01 had completed a full migration away from VMware, according to CloudBolt’s own report. It is the only completion figure traced to a primary source on this page. Broadcom’s infrastructure software segment, which includes VMware, CA and Symantec, reported net revenue of $27,029 million ^S15 for the 2025 fiscal year in its Form 10-K. That is up 26% ^S15 on the $21,478 million ^S16 in its fiscal 2024 Form 10-K. Public surveys of VMware customers since Broadcom closed the acquisition on 2023-11-22 are collected below, each figure shown with its wording, time bound, sample, fielding period and sponsor.

Key findings

• 86% ^S01 of surveyed VMware decision makers (CloudBolt-sponsored, n=302, North America) are actively reducing their VMware footprint; the measure has no time bound (January 2026).
• 4% ^S01 of those surveyed decision makers have completed a full migration away from VMware, according to CloudBolt’s report on its own survey (January 2026). It is the only completion figure traced to a primary source on this page.
• 56% ^S07 planned to decrease overall VMware usage within the next 12 months (sponsor Veeam, research by Foundry, n=561, base: all respondents) (dated: fielded June to July 2024).
• 43% ^S13 were very or extremely likely to switch hypervisor platform within 3 years, in a sample of all virtualization users (Red Hat, research by Illuminas, March to April 2025, n=1,010).
• 90% ^S11 are considering alternatives to VMware due to higher licensing costs (Rimini Street, research by Unisphere, December 2025 to February 2026, n=269).
• 59% ^S01 experienced VMware cost increases of more than 25% since the acquisition, according to CloudBolt’s report (January 2026, base: all respondents).
• 48% ^S04 of Civo’s 1,000 ^S04 current or recent VMware users already pay more for the same features (sponsor Civo, fieldwork not stated) (dated: published July 2024).
• 72% ^S01 of migrating workloads are heading to public cloud IaaS (CloudBolt, January 2026, multi-select, base: migrating workloads).
• 91% ^S09 planned to stay with VMware in a survey by OVHcloud US, a VMware Cloud Service Provider (n not disclosed, fieldwork not stated, published 2024-10-09) (dated: fielded before October 2024).
• $12,384 million ^S16 is the only VMware-only revenue figure Broadcom has filed, for about 49 weeks of the 2024 fiscal year. That is 57.7% ^derived of the segment’s $21,478 million ^S16 that year.
• 20% ^S05 of VMware enterprise customers would begin to leave the stack in 2024, in Forrester’s prediction; a forecast, not a survey result (dated: published October 2023).

Every survey at a glance

Every wave with a publishable figure is listed, plus one pre-close baseline with no figures used. Fielding periods are as stated by the publisher unless marked inferred or not stated. Dated marks follow the cut-off in the Methodology section below. When fieldwork is not stated, the publication date is used. Veeam 2026 is held back because its report is form-gated and was not retrieved. Everpure 2025 is held back because it states no fielding date and its headline conflicts with its own chart.

A vendor selling alternatives, migration tooling or third-party support has an interest in a leaving result; a VMware reseller has an interest in a staying result. No neutral, publicly documented survey of VMware customers was found.

Intent to reduce or leave VMware

Three publishable rows measure intent to reduce VMware usage, and each asks a different question. The chart sets them side by side with question wording, time bound and base under each bar. It is a reference, not a trend.

Veeam’s survey of 561 ^S07 senior IT decision makers (research by Foundry) covered companies with 1,000+ employees in France, Germany, the UK and the US. It reported the share “planning to decrease their overall VMware usage within the next 12 months”. That share was 56% ^S07 (dated: fielded June to July 2024).

Red Hat’s survey of 1,010 ^S13 IT decision makers and virtualization practitioners (research by Illuminas) was fielded March to April 2025 in the US, UK and English-speaking APAC. Respondents were all virtualization users, not VMware customers only. The measure was platform-agnostic. In the report body, the measure is how likely they were “to switch to a different hypervisor platform in the next 3 years”. Its callout reads “very/extremely likely to switch to a different or additional hypervisor”. 43% ^S13 were very or extremely likely and 33% ^S13 somewhat likely. Combined, 76% ^derived reported some likelihood of switching, the sum of the two Red Hat figures.

CloudBolt sponsored a survey of 302 ^S01 IT decision-makers at director level and above, fielded in January 2026 by an unnamed independent research firm, at North American companies with 1,000+ employees. The measure had no time bound and counted any reduction: respondents who “report they are actively reducing their VMware footprint”. That share was 86% ^S01 of those surveyed VMware decision makers. CloudBolt’s report adds that 4% ^S01 had completed a full migration away.

CloudBolt’s report also says 25% ^S01 of its January 2026 respondents rated the acquisition extremely or very disruptive. Its earlier wave, run by Wakefield Research, put that share at 46% ^S02 (dated: fielded May 2024). The populations differ. The 2026 wave covered director-level decision-makers at North American companies with 1,000+ employees. The 2024 wave covered enterprise IT decision-makers at companies of all sizes, region not stated. Both waves asked about disruption at the time of the survey, and both bases are all respondents.

Two Rimini Street waves measure intent to evaluate, not intent to reduce, and are kept off the chart. 98% ^S10 were already using, planning to use or considering alternatives for at least part of their VMware estate (DBTA/Unisphere Research, n=110+) (dated: fielded Q3 2024). 90% ^S11 were considering alternatives to VMware due to higher licensing costs (Unisphere Research, December 2025 to February 2026, n=269).

Everpure’s 2025 report also asked about reducing VMware. Its percentages are held back; see How to read these numbers.

Why customers are moving away from VMware

Two sponsors published reasons for looking at alternatives: Rimini Street, which sells third-party support, and CloudBolt, which sells cost management for both paths. The table is grouped by survey wave, and each row states its own sponsor, fieldwork, base and time bound. Reasons are not exclusive, so shares within one wave add up to more than the whole sample. Shares from different waves answer different questions and are not ranked against each other.

Rimini Street’s newer wave covered 269 ^S11 executives, managers and professionals at VMware user organisations across a global sample. Its earlier wave covered 110+ ^S10 VMware customers (dated: fielded Q3 2024). CloudBolt’s earlier wave asked 300 ^S02 enterprise IT decision-makers for their top reasons for considering a move away from VMware in part or in whole (dated: fielded May 2024). In that wave, three reasons tied at the top.

Price increases

Every percentage here states its base. A share of all respondents and a share of a subset are not comparable.

Civo surveyed 1,000 ^S04 current or recent VMware users and did not state the fielding date (dated: published July 2024). Nearly 70% ^S04 expected a price increase (base: all respondents). 48% ^S04 already paid more for the same features since the acquisition (base: all respondents). Among those who faced increases, 48% ^S04 said costs had doubled, 30% ^S04 reported a fourfold increase and 15% ^S04 tenfold. The base for those three is the 48% who faced increases, not all 1,000 respondents. Civo does not say whether the three multiples overlap. This page reads them as separate outcomes, as Civo words them: doubled, fourfold and tenfold. Re-based to all respondents, 23.0% ^derived of Civo’s sample saw their costs double: 48% ^S04 paying more multiplied by 48% ^S04 doubled.

CloudBolt’s two waves separate expectation from experience. 73% ^S01 of the earlier wave expected VMware costs to more than double (Wakefield Research, n=300, base: all respondents, restated in the 2026 report) (dated: fielded May 2024). 59% ^S01 of the January 2026 wave had experienced increases of more than 25% since the acquisition, according to CloudBolt’s report (January 2026, n=302, base: all respondents).

92% ^S10 of Rimini Street’s respondents anticipated additional VMware price increases within the next 12 to 18 months (DBTA/Unisphere Research, n=110+, base: all respondents) (dated: fielded Q3 2024).

Where migrating workloads are going

Destination questions differ in base: CloudBolt 2026 asks about migrating workloads, Veeam and CloudBolt 2024 ask all respondents, and OVHcloud asks on-premises users only. All are multi-select or non-exclusive, so shares within one survey add up to more than the whole sample. Only CloudBolt 2026 states a time bound, current migration plans; the other three waves state none.

72% ^S01 of migrating workloads were heading to public cloud IaaS in CloudBolt’s survey (fielded January 2026, n=302, base: migrating workloads, time bound: current migration plans, multi-select). CloudBolt’s Hyper-V/Azure Stack share is withheld because its press release and report page give different values. The SaaS share matches in both documents, but CloudBolt states it in the same line as the Hyper-V/Azure Stack share, and that line is held until CloudBolt confirms the conflicting figure.

Veeam’s respondents named both paths. 39% ^S07 would migrate some virtual machines to other hypervisors in their own data centres. 36% ^S07 would move some VMs to cloud-based alternatives (Foundry, fielded June to July 2024, n=561, base: all respondents, no time bound stated) (dated: fielded June to July 2024).

CloudBolt’s 2024 wave asked for likely go-forward scenarios (Wakefield Research, about May 2024, n=300, base: all respondents, multi-select, no time bound stated) (dated: fielded May 2024). Staying with VMware fully was chosen by 40% ^S02 and staying partially by 43% ^S02. Moving more workloads to public cloud was chosen by 38% ^S02, migrating to an alternative hypervisor by 34% ^S02 and migrating entirely to public cloud by 33% ^S02.

OVHcloud US, a VMware Cloud Service Provider that sells managed VMware, reported that 62% ^S09 of respondents using VMware on-premises expected to move to a cloud or hybrid cloud solution. 38% ^S09 planned to continue on-premises (base: respondents using VMware on-premises, n not disclosed, no time bound stated, fieldwork not stated, published 2024-10-09) (dated: fielded before October 2024).

Broadcom’s numbers

VMware stockholders received about $30,788 million ^S16 in cash, plus Broadcom shares worth $53,398 million ^S16 at fair value. Broadcom closed the acquisition on 2023-11-22. Broadcom reports VMware inside its infrastructure software segment, together with CA Technologies mainframe software and Symantec enterprise security. The fiscal year ends on the Sunday nearest 31 October. All values are US dollar millions as filed.

Segment net revenue was $7,637 million ^S16 in the 2023 fiscal year, the last full year before VMware (dated: fiscal year ended October 2023). It rose to $21,478 million ^S16 in the 2024 fiscal year, a 53-week year with about 49 weeks of VMware ownership. The 181% ^S16 increase on fiscal 2023 is the acquisition itself, not organic growth. The segment reported $27,029 million ^S15 in the 2025 fiscal year. That is up 26% ^S15 in the first full year of VMware ownership, and the segment was 42% ^S15 of Broadcom’s total net revenue.

The only VMware-only revenue line in any Broadcom filing is $12,384 million ^S16 of net revenue attributable to VMware for the 2024 fiscal year, about 49 weeks of ownership. That is 57.7% ^derived of the segment’s net revenue for that year: $12,384 million ^S16 divided by $21,478 million ^S16. Because VMware was owned for about 49 of 53 weeks, the full-year share would be higher. No VMware-only revenue figure exists for the 2025 fiscal year, and Broadcom files no churn or customer-count metric.

In the quarter ended 2025-11-02, the segment, which also holds CA and Symantec, reported $6,943 million ^S17, up 19% ^S17 on $5,824 million ^S17 a year earlier. In the quarter ended 2026-02-01, the segment, with CA and Symantec, reported $6,796 million ^S18, up 1% ^S18 on $6,704 million ^S18. In the quarter ended 2026-05-03, the segment, with CA and Symantec, reported $7,178 million ^S22, up 9% ^S22 on $6,596 million ^S22. In the quarter ended 2026-08-02, the segment, with CA and Symantec, reported $8,752 million ^S23, up 29% ^S23 on $6,786 million ^S23.

Analyst forecasts

Analyst forecasts are predictions, not measurements, and sit apart from the survey figures. Forrester predicted that 20% ^S05 of VMware enterprise customers would begin to leave the VMware stack during 2024, in its Predictions 2024 blog, before the acquisition closed (dated: published October 2023). Gartner forecasts on VMware workload migration exist only as press reports of conference remarks or Magic Quadrant text, with years and baselines that differ between outlets. None appears on this page.

How to read these numbers

Who paid. Every survey here was commissioned by a company with a commercial interest in the answer. Civo, Red Hat and VergeIO sell alternatives to VMware. Rimini Street sells third-party support for perpetual licences. Veeam sells backup and markets migration tooling. OVHcloud US resells managed VMware and is the only sponsor with an interest in customers staying. CloudBolt sells cost management for both paths.

Time bounds. Intent to reduce is a different question in every survey. CloudBolt 2026 counts any current reduction with no time bound. Veeam 2024 asks about the next 12 months. Red Hat 2025 asks all virtualization users about switching hypervisor platform in the next 3 years. Rimini Street asks about considering alternatives, which is intent to evaluate. A question with no time bound captures everyone a bounded question captures, plus those who plan to act later, so the two are not comparable.

Bases. Civo’s price multiples are shares of the 48% ^S04 who faced increases, not of all 1,000 ^S04 respondents (dated: published July 2024). CloudBolt’s percentages are of all respondents, except the destination question, which is of migrating workloads.

Completion. 4% ^S01 completed a full migration away, according to CloudBolt’s report on its January 2026 survey. It is the only completion figure traced to a primary source on this page. Other numbers quoted online as already switched are intent figures, or appear only in news coverage of gated reports, and are not used here.

Fielding date, not publication date. Each wave is dated by when it was fielded; Rimini Street’s 2026 report was published on 2026-10-06 but fielded between December 2025 and February 2026. The dating rule and its cut-off date are in the Methodology section below. Dated figures are a refresh trigger.

Unverified documents. CloudBolt’s 2026 press release and report page disagree on several statistics; conflicting figures are not shown. Everpure’s 2025 report states no fielding date and its headline conflicts with its own chart. Veeam’s 2026 report is form-gated and was not retrieved. Virtified’s 2026 survey of IT professionals is excluded: its figures appear only in a syndicated copy of its press release, not on virtified.com, and none is shown here. Gartner forecasts and Broadcom CEO statements exist only as press reports. None of these figures appears here.

Methodology

Every figure traces to a line in the source ledger and to a row in the clean dataset. Sources were retrieved on 2026-10-07, except the Broadcom Q2 and Q3 FY2026 releases ^{S22 S23}, retrieved on 2026-10-19. The last data check was on 2026-10-19. Survey percentages are reproduced exactly as published. Derived figures are rounded to one decimal place and state their formula where used. Charts and the downloadable dataset carry publishable rows only. Hub-wide rules are on the methodology page.

Dating rule and cut-off. A figure is labelled dated when the date that counts for it is before 19 October 2024, which is 24 months before the last data check. For a survey, that date is the end of fieldwork, or the publication date when fieldwork is not stated. A wave published on the cut-off day counts as fielded before it. For a Broadcom filing, it is the end of the period the filing reports. For a forecast, it is the publication date. The date of an event, such as the closing of the acquisition on 2023-11-22, never makes a figure dated. A derived figure is dated when any of its inputs is dated. Dated figures are a refresh trigger.