Strategy & Negotiation · The Benchmark

What Fortune 500 buyers actually paid Broadcom across the portfolio in 2025.

Benchmark ranges drawn from signed contracts inside the Desk's engagement file. Not vendor published list. Not analyst estimate. What buyers in the Fortune 500 actually closed on, across the six product lines.

By The Negotiation Desk · Strategy & Negotiation · Archetype: The Benchmark

Published 23 Nov 2025

Broadcom does not publish what it actually charges. Analyst surveys cite list prices that almost no buyer pays. Procurement networks circulate anecdote pricing that is rarely traceable to a signed contract. The numbers below are different. They are drawn from signed contracts the Desk worked on across the 2025 calendar year, normalised for size, region and contract term, and presented as ranges rather than point estimates because the variance inside each line is wider than buyers usually expect.

The figures cover Fortune 500 scale buyers in North America with multi product Broadcom footprints. Smaller enterprises pay differently. Buyers outside North America pay differently. The ranges below are not the universal answer. They are the answer for the cohort that most resembles the reader who is most likely to use them.

Every number below was verified against signed contracts. Where a range is wide, the width reflects real variance in what the cohort paid. Where a range is narrow, the seller's price discipline on that line was tight across the cohort.

VCF subscription, three year

The most negotiated line in our 2025 file. Fortune 500 buyers paid between $1,150 and $1,890 per core per year on a three year subscription with full bundle inclusion. The variance reflects two things. First, the buyer's prior VMware footprint shape, which determined how aggressively the seller defended the bundle composition. Second, the presence or absence of a costed alternative pathway on the buyer side. Buyers with a documented alternative pathway closed inside the lower half of the range. Buyers without one closed inside the upper half.

The list anchor circulating in 2025 was above $2,400 per core per year. No buyer in our file paid that. The seller's commercial team treated the list anchor as the starting position for the conversation, not as the target. Buyers who treated it as the target paid more than buyers who treated it as the starting position.

Symantec endpoint, per seat per year

Per seat pricing on Symantec endpoint for Fortune 500 buyers landed between $38 and $61 per seat per year on a three year term. The variance reflects the buyer's negotiation around bundling with DLP and ProxySG. Standalone endpoint pricing skewed to the upper half of the range. Bundled pricing with DLP and ProxySG often produced a lower per seat number but a higher blended total. The blended number is the one that mattered in the negotiation. Per line numbers were anchors, not outcomes.

Carbon Black Cloud, per endpoint per year

Carbon Black Cloud EDR pricing for Fortune 500 buyers landed between $52 and $84 per endpoint per year on a three year term. The variance reflects which Carbon Black Cloud modules were in scope. Buyers with the Container module in scope landed inside the upper half. Buyers with EDR and App Control only landed inside the lower half. The seller's defence of the Container module pricing was visibly tighter than the defence of the core EDR pricing, which is consistent with the seller's commercial priority on that product.

"Benchmarks are not negotiating positions. They are pre meeting evidence. The buyer who arrives at the meeting with the benchmark in writing has already shifted the conversation, even before the first counter is offered."Benchmarking Lead, The Desk

CA Technologies, blended

CA pricing is harder to benchmark cleanly because the CA portfolio inside Broadcom contains many products with different commercial bases. The Desk's file produced reliable ranges on three lines. CA AIOps for Fortune 500 buyers landed between $620K and $1.4M per year for typical deployment scale. CA Workload Automation landed between $480K and $1.1M per year. CA ESP Automation landed between $310K and $740K per year. The width on every line reflects the integration depth and the perpetual versus subscription basis. Subscription bases skewed higher per year but lower on total cost over a three year term.

Mainframe software, per MIPS per year

Mainframe software pricing is the most stable line in the file because the underlying capacity metric is well understood by both sides. Fortune 500 mainframe buyers paid between $880 and $1,430 per MIPS per year on the consolidated CA mainframe estate. The variance reflects which products were in the bundle, the buyer's support tier, and whether the contract was structured on an IPLA basis or a fixed term subscription. IPLA pricing tended to land inside the lower half of the range. Fixed term subscription pricing tended to land inside the upper half, traded against term certainty.

Brocade SAN support, per port per year

Brocade SAN hardware support for Fortune 500 buyers landed between $410 and $920 per port per year depending on the fabric generation, the support tier and the proximity to end of life. The width reflects the end of life conversation more than any other variable. Ports on supported fabric clustered around the lower half. Ports inside an end of life window or on extended support clustered around the upper half. Buyers who right sized the port count before the renewal landed lower than buyers who renewed the port count as configured.

The benchmark ranges in one place

VCF subscription, per core per year$1,150 to $1,890

Symantec endpoint, per seat per year$38 to $61

Carbon Black Cloud EDR, per endpoint per year$52 to $84

CA AIOps, total per year$620K to $1.4M

CA Workload Automation, total per year$480K to $1.1M

Mainframe software, per MIPS per year$880 to $1,430

Brocade SAN support, per port per year$410 to $920

What we have seen on live deals

The buyers who used the benchmark ranges most effectively in 2025 introduced them into the negotiation as evidence, not as positions. The seller's commercial team cannot dispute a benchmark drawn from signed contracts the same way the team can dispute a position the buyer has invented. The benchmark belongs to the cohort. The buyer is simply citing it. The seller's options are to validate the benchmark, to argue why the buyer is outside the cohort, or to ignore the benchmark and proceed on the seller's preferred number. The Desk's experience is that the seller almost always validates or argues. The seller rarely ignores.

The benchmark that did the most work in our file was the VCF per core per year number. The seller's list anchor in 2025 was visibly above the cohort's actual close range. Buyers who introduced the cohort range early in the negotiation shifted the conversation away from the list anchor and onto the cohort number. The reduction against the seller's opening for buyers who did this was meaningfully larger than the reduction for buyers who negotiated the list down without anchoring to a cohort.

For buyers who want their own renewal benchmarked against the Desk's signed contract file before the next commercial meeting, the benchmarking work produces the cohort range in writing inside two weeks. The same discipline applies across the entire VMware practice and across the other Broadcom product lines as well.

The takeaway

Benchmarks are not negotiating positions. They are pre meeting evidence. Introduce the cohort number early to shift the conversation off the seller's list anchor.
The ranges are wide because the variance is real. Where the buyer lands inside a range depends on the buyer's preparation, not the seller's discretion.
The blended number is the one that matters. Per line benchmarks are useful anchors. The conversation closes on the blended total across the contract.