The 2026 Enterprise Software Overpayment Index
The 2026 Enterprise Software Overpayment Index measures the gap between vendor first-proposal pricing and signed-contract pricing across 213 enterprise software renewals. Aggregated by vendor, the median first-proposal overpayment in Y1 of an enterprise software contract is 28%; by Y3 it compounds to 47%. Oracle (38% Y1), Broadcom/VMware (36%), and SAP RISE (33%) lead the index. This page publishes the medians, the methodology, and the per-vendor table, drawn entirely from NoSaveNoPay negotiated engagements between January 2024 and May 2026.
Headline findings
Three findings drive the 2026 Index. First, the gap between what enterprise vendors propose and what they sign is wider than buyers expect, and it has widened versus the 2023 baseline we tracked internally. Second, the gap is not random across vendors: Oracle, Broadcom-era VMware, and SAP RISE consistently produce the largest first-proposal overpayments, while AWS EDP and Google Cloud commitments produce the smallest. Third, escalators matter more than headline discounts. In every vendor we measured, Y3 overpayment exceeded Y1 overpayment by 50% or more, because vendor proposals embed annual increases of 5-12% that compound. The full methodology is documented below; the per-vendor table is the dataset.
The per-vendor table, median Y1, Y2, Y3 overpayment
Each row is the median overpayment percentage across all NoSaveNoPay engagements for that vendor between January 2024 and May 2026. Overpayment is defined as (baseline − signed value) / baseline, where baseline is the vendor's first written proposal and signed value is the executed contract value, normalised per year for Y1/Y2/Y3 cuts. Vendors with fewer than 7 engagements are excluded.
| Vendor | N | Y1 overpayment | Y2 overpayment | Y3 overpayment |
|---|---|---|---|---|
| Oracle | 31 | 38% | 45% | 53% |
| Broadcom / VMware | 12 | 36% | 44% | 51% |
| SAP (RISE) | 19 | 33% | 41% | 49% |
| IBM | 14 | 31% | 38% | 46% |
| Microsoft EA | 42 | 28% | 35% | 43% |
| Salesforce | 22 | 27% | 34% | 42% |
| ServiceNow | 8 | 26% | 33% | 40% |
| Workday | 8 | 24% | 31% | 38% |
| Top-15 SaaS (mixed) | 18 | 23% | 29% | 36% |
| AWS EDP | 21 | 19% | 24% | 29% |
| Google Cloud commit | 9 | 18% | 23% | 28% |
| Index median (all vendors) | 28% | 35% | 42% | |
| Index mean (weighted by deal value) | 29% | 37% | 47% | |
N = number of engagements per vendor in the sample. Sample window 2024-01-01 to 2026-05-18. Methodology footnoted below. Underlying data confidential under client MSAs; aggregate medians released under NSP terms.
Why Oracle leads the index
Oracle's 38% Y1 median is the highest in the dataset for three structural reasons. First, Oracle Unlimited Licence Agreement renewals are routinely proposed at "list price holding", meaning the vendor offers to renew at the same Unit Cost Reference Model price but on an inflated metric base, producing a first-proposal that ignores three years of right-sizing opportunity. Second, Oracle's first proposals frequently include support escalators of 8% annually, well above CPI, which the buyer has both the leverage and the precedent to push back to 3-5%. Third, Oracle ULA exits and certifications create one-off negotiation leverage that buyers leave on the table without specialised support. The Oracle Negotiation service page covers the playbook in detail. For an example of how this plays out in practice on a real ULA renewal, see the Fortune 500 Bank case study.
Why Broadcom/VMware climbed the index
The 36% Y1 median for Broadcom-era VMware reflects post-acquisition pricing reset. Since the November 2023 Broadcom acquisition, VMware ELA renewals are issued under the new portfolio bundle (vSphere Foundation, Cloud Foundation) at 2-4× the previous per-socket cost, with subscription replacing perpetual licensing. The first-proposal premium relative to a realistic re-negotiated outcome, incorporating workload-level licensing right-sizing, partial migration to alternative hypervisors as leverage, and credible exit threats, averages 36% across our 12 Broadcom/VMware engagements. The Broadcom/VMware Negotiation service page documents the standard playbook; the Global Manufacturer case study shows it executed on a $14M renewal.
Why AWS and Google Cloud sit at the bottom of the index
Hyperscaler commitment pricing (AWS EDP, Google Cloud commit, Azure MACC) shows the lowest overpayments, 18-19% Y1 medians, for two reasons. First, hyperscaler pricing is more transparent than on-premise software pricing: list rates are public, committed-spend discounts follow disclosed bands, and competitive bids from peer hyperscalers are credible. Second, AWS and Google have stronger competitive constraints than Oracle or SAP in the renewal moment. Our AWS Negotiation and Google Cloud Negotiation service pages cover the tactical playbook; the Energy Company AWS EDP case study documents a 21% effective discount delta on a $14M commit.
The escalator multiplier, why Y3 is worse than Y1
Across every vendor in the index, Y3 overpayment exceeds Y1 overpayment by 50-60%. This is the escalator multiplier. Vendor first proposals embed annual increases of 5-12% for on-premise enterprise software, 4-8% for SaaS, and 3-7% for hyperscaler commits. A first-proposal Y1 overpayment of 28% with a 7% annual escalator compounds to a 47% Y3 overpayment relative to a negotiated path where escalators are capped at CPI plus 2%. Reducing escalators is the single highest-leverage move in any multi-year enterprise software renewal, see SaaS price increase response for the SaaS-specific framing.
Methodology
Sample frame
All NoSaveNoPay-led enterprise software and cloud negotiations completed between 2024-01-01 and 2026-05-18. N=213 across 11 vendor categories. Excludes engagements where NSP was advisory-only (no negotiation execution) or where the buyer signed before NSP's playbook was applied.
Baseline definition
The vendor's first written proposal for the renewal or new purchase, on vendor letterhead, dated. Not a benchmark, not a database value, not a number NSP invents. The same document the buyer would have used to authorise the renewal in-house.
Signed value
Total commercial value of the executed agreement over the full contract term, on the same basis as the baseline (same scope, same term, same inclusions). All values normalised to USD at the signature-date FX rate.
Annualisation
Y1/Y2/Y3 medians are derived by allocating multi-year total contract values to individual years using the explicit annual schedule in the contract. Where escalators are stepped, the actual stepped value is used; where escalators are blended into a single TCV, the implied annual schedule is reconstructed from the vendor's first proposal.
Exclusion rules
Excluded from the dataset: engagements where the buyer added scope mid-negotiation that was not in the baseline; engagements where vendor M&A activity (e.g., acquisition during the renewal) created a non-comparable basis; engagements under NDA that prohibits aggregation; vendors with fewer than 7 engagements in the window.
Statistical caveats
Medians reported rather than means to reduce sensitivity to outlier engagements (e.g., the single largest Microsoft EA at $18.3M baseline → $13.1M signed). The mean weighted by deal value sits 1-5 percentage points above the median for every vendor, indicating larger deals see slightly higher first-proposal premiums.
How NoSaveNoPay closes the gap
The Index documents the gap; it does not close it. Closure runs through a negotiation playbook that varies by vendor but follows the same five-step NoSaveNoPay process: usage diligence, baseline countersignature, leverage map, vendor engagement, and signed contract with CFO-verified savings memo. Our fee is 25% of verified savings; the buyer keeps 75%. If we don't save anything, you don't owe anything, see the full gainshare model or the pricing page for commercial detail. For the engagement contract language itself, the gainshare engagement letter template walks through every clause.
How to cite the Index
NSP Overpayment Index 2026. NoSaveNoPay LLC, May 2026. Available at https://nosavenopay.com/research-2026-overpayment-index. Researchers and journalists requesting custom anonymised cuts (industry vertical, deal-size band, geography) can contact research@nosavenopay.com, turnaround is typically 5 business days. The 2027 edition will publish in Q2 2027 with three additional vendors (Snowflake, Databricks, Anaplan).
Related research and reading
- Vendor Price-Increase Tracker 2026, quarterly-updated list-price changes by vendor.
- 2026 Software Audit Likelihood Index, by-vendor probability of audit per enterprise per year.
- The Gainshare Software Negotiation Model, pillar reference.
- How gainshare fees are calculated, the math behind 25% of verified savings.
- CFO guide to software spend, boardroom briefing format.
- Case studies, eleven worked examples of the playbook in execution.
Where does your contract sit on the Index?
Send us the vendor name, the contract type, and the renewal window. We will return a free, written estimate of the first-proposal overpayment for your situation, using the same methodology behind the Index. No commitment. No fee until savings are CFO-countersigned.
Get a Free Estimate → Use the CalculatorFrequently asked questions
How is overpayment defined in this Index?
Overpayment is the percentage difference between the vendor's first written proposal for a renewal or new purchase (the baseline) and the signed contract value, expressed as a percentage of baseline. If the baseline was $10M and the signed value was $7M, overpayment in the first proposal was 30%. The Index reports the median across all engagements per vendor for Y1, Y2 and Y3 of the engagement term.
What is the sample size?
The Index aggregates 213 enterprise software negotiations completed by NoSaveNoPay between January 2024 and May 2026. Per-vendor sample sizes range from 8 (Workday, ServiceNow) to 42 (Microsoft). Vendors with fewer than 7 engagements are excluded from the published medians.
Which vendor has the highest baseline overpayment?
Oracle has the highest median Y1 baseline overpayment at 38%, driven by aggressive ULA renewal proposals and historical price escalators that are routinely negotiated down. Broadcom-era VMware ranks second at 36% on year-one ELA proposals issued since the 2023 acquisition. SAP RISE is third at 33%.
Does overpayment grow over time?
Yes, Y2 and Y3 baseline overpayment is materially higher than Y1 in every vendor we measured, because vendor proposals typically embed annual escalators of 5-12%. By Y3, the cumulative overpayment versus a negotiated path averages 47% across the index, compared to 28% in Y1.
Why is the Index focused on first-proposal versus signed-value?
This is the only auditable, vendor-issued benchmark. Industry surveys rely on self-reported pricing data, which is biased toward firms willing to disclose. The vendor's first proposal is a contemporaneous document on vendor letterhead, the same document the buyer would have used to authorise the renewal in-house. The signed contract is the achieved outcome. The difference is the overpayment that would have occurred without negotiation.
Can I download the raw data?
The aggregated per-vendor table on this page is the public release. The underlying engagement-level data is confidential under each client's MSA. Researchers and journalists can request additional cuts (industry vertical, deal-size band, region) by contacting research@nosavenopay.com, we can prepare custom anonymised summaries within 5 business days.
How does this compare to Gartner or Forrester pricing data?
Gartner and Forrester surveys typically ask buyers to report final negotiated pricing relative to list price. They do not capture the first-proposal-to-signed-value differential, which is the negotiation-yield number. Our Index complements those datasets by showing how much room exists between vendor opening offers and achievable outcomes.
Will the Index be updated?
Annually. The 2027 edition will include Q3 2026 through Q2 2027 engagements and add three vendors currently below the 7-engagement threshold (Snowflake, Databricks, Anaplan). Quarterly updates on individual vendor medians are available via the Vendor Price-Increase Tracker.