Every major enterprise software price increase from 2019 to 2026 belongs on one timeline. This report places those moves in order, by vendor, with the trigger that preceded each one and the realized cost the buyer paid.
Every major enterprise software price move since 2019 sat on a timeline that was visible months ahead. This report places those moves in order, vendor by vendor, with the trigger that preceded each one and the realized cost the buyer paid.
About this timeline
The Redress Compliance vendor price hike timeline is a directional reference, not a price list. It draws on three inputs.
We report bands and directions, not precise discounts. Individual outcomes vary widely with estate size, timing, and leverage. Where a single number appears, treat it as the middle of a defensible range.
The pattern is simple, and it has held in every major case on this timeline. A price increase follows a structural change that the vendor has already signaled. The structural change is usually one of three things, and it is almost always public months before the rate change lands.
The three triggers we see, in order of how reliably they predict a move, are an ownership change at the vendor, a licensing model or metric change, and the launch of a generative AI add on tied to the existing product.
A new owner almost always reprices. Private equity, strategic acquirers, and activist driven boards run a similar playbook. They look at the installed base, identify the underpriced cohort, and reset.
The Broadcom acquisition of VMware is the cleanest example, but the same dynamic showed up in smaller form across IBM and Red Hat, and across several mid market roll ups in the security and observability space during the period.
A change to the licensing metric is a price change in disguise. Oracle Java moved from the processor metric to the employee metric in 2023. The list rate did not change, but the multiplier did, and the bill rose for almost every estate that was previously sized on Java footprint alone.
SAP did a softer version of the same move with indirect access. The metric clarified, the cost shifted, and the buyer who had not modeled the new measure paid first.
Every generative AI launch on this timeline carried a price tag attached to the existing seat. Microsoft 365 Copilot, Salesforce Einstein and then Agentforce, ServiceNow Now Assist, and Adobe Firefly each added a recurring per seat or consumption charge on top of the base license.
The AI launch is the most reliable signal of the next renewal increase, because the add on margin is what funds the vendor through the slower base growth that typically follows a model reset.
Oracle reset the Java commercial model twice in the period, and each reset materially raised the effective cost for almost every customer that touched commercial Java.
In April 2019, Oracle moved commercial Java from the paid update model to a named subscription, the Java SE Subscription. Free public updates for Java 8 ended for commercial use. Customers who had treated Java as effectively free for years now needed a paid contract to receive security patches on the runtime they were already running in production.
The subscription was priced per processor for servers and per user for desktops. For most estates the move added a recurring annual line that had not existed before, and the first contract was almost always larger than expected.
In January 2023, Oracle replaced the user and processor metrics with a single metric counted across the whole company, the Java SE Universal Subscription. Every employee, contractor, agent, and consultant of the customer is in scope, regardless of who actually runs Java.
For a customer that had a few hundred Java users on a few servers, the bill could jump by a factor of five to ten or more under the new metric. Public estimates from analyst houses placed typical increases in the two to five times band for mid sized estates and far higher for large estates with small Java footprints.
The realized outcome in our engagement file diverged sharply by buyer posture. Customers who modeled the new metric early, removed Oracle Java where possible, and presented a credible OpenJDK alternative settled at a fraction of the opening ask. Customers who responded after the first quote often signed the headline number.
Through 2024 and 2025, Oracle leaned heavily on the audit motion to drive Java subscription conversions. Customers received Java usage queries from Oracle License Management Services on a cadence that matched renewal cycles, and the line between an audit and a renewal blurred for many estates.
The buyer side lever was preparation, not protest. An estate that had already documented its Java footprint, isolated the commercial use cases, and stood up an OpenJDK migration plan walked into the conversation with the leverage the timeline implied.
The Java move did not exist in isolation. Oracle ran a parallel push on Unlimited License Agreement exits and database renewals, and the leverage on one negotiation often spilled into the other.
Buyers who treated Java as a standalone problem missed the connection. The reliable approach was to map the entire Oracle footprint, identify the negotiations that would converge, and sequence the renewal calendar accordingly.
Microsoft 365 had not raised commercial list prices for a decade before 2022. Two list moves followed in quick succession, and the Copilot add on layered a much larger cost on top of the base seat.
In March 2022, Microsoft raised commercial Microsoft 365 and Office 365 list prices for the first time since the suites launched. The increases ran from roughly nine to twenty five percent by SKU. Microsoft published the new rate card alongside its broader commercial plan comparison shortly after.
The renewal impact was uneven. Customers on a Microsoft Enterprise Agreement could lock the old price for the remaining term, while customers on monthly subscriptions or CSP saw the new price almost immediately. The lever was contract type, not vendor concession.
Microsoft announced general availability of Microsoft 365 Copilot at 30 dollars per user per month in November 2023. For a Microsoft 365 E3 customer at around 36 dollars per user per month, the Copilot add on roughly doubles the cost of a knowledge worker seat.
The add on is the largest single change in Microsoft 365 economics since the suite launched. Even partial attach at a fraction of seats reframes the renewal conversation, because the base seat now sits inside a bundle conversation rather than a standalone one.
In early 2026, Microsoft raised commercial prices again on selected SKUs, with the move concentrated on Microsoft 365 E3 and E5 and on certain security and compliance add ons. The realized impact for a buyer in mid renewal landed in a range that depended almost entirely on the price lock language in the existing Enterprise Agreement.
The Microsoft Enterprise Agreement renewal cycle is now the central pricing event for most large Microsoft customers. The True Up motion, the SCE versus EA choice, and the inclusion or exclusion of Copilot in the base bundle all carry material cost implications.
The reliable buyer side approach to the Microsoft EA in 2026 is to separate the SKU count discussion from the AI attach discussion, and to negotiate the True Up rules and price hold language with the same rigor as the headline seat rate.
The reliable move in 2026 is to model the Copilot add on as a separate contract rather than a bundle line, even when the vendor prefers a bundle. Separating the term, the cap, and the exit on the AI add on protects against a second repricing inside the contract.
The Microsoft pricing story in 2026 is no longer only about Microsoft 365. Azure AI services, OpenAI on Azure, and the inference cost on Copilot Studio agents now sit alongside the seat price as material budget lines for many buyers.
The reliable approach is to model the Azure AI consumption line on its own commit and rate card, and to keep the negotiation separate from both the seat and the Copilot add on.
The Broadcom acquisition of VMware closed on November 22, 2023. The pricing reset followed in December 2023 and January 2024, and it was the steepest single repricing on this timeline.
Broadcom retired most of the standalone VMware SKUs and consolidated the portfolio into a small number of bundles, with VMware Cloud Foundation and VMware vSphere Foundation at the top of the stack. Perpetual licenses were removed for new sales. Existing perpetual customers were directed to renew on a subscription term.
The conversion math is the entire story. A customer paying a small annual support fee on a fully amortized perpetual estate now faced a full subscription fee on the bundle, and the bundle included SKUs the customer may not have previously licensed.
In our engagement file, transition quotes from perpetual to the new bundles commonly landed in the 100 percent plus uplift band on annualized cost. A smaller cohort of public cases ran into multiples of the prior annual support, particularly where the buyer had bought standalone vSphere and now needed to take the full Foundation bundle.
The realized outcome split sharply. Buyers who negotiated term length, declared which bundle SKUs they would actually deploy, and ran a parallel evaluation of Nutanix, Proxmox, or Microsoft Hyper V often held the increase below the opening ask by a wide margin.
The Broadcom reset was the clearest case on the timeline where the lever was a credible alternative, not a vendor concession. Buyers who had a documented exit path on at least part of the estate by mid 2024 saw materially lower quotes than buyers who did not.
The post reset VMware customer base split into three cohorts. The first signed the new bundle terms at the headline rate and absorbed the increase as a strategic choice. The second negotiated the bundle and term aggressively, often holding the increase below the opening ask by a wide margin.
The third cohort started an exit. Partial migration to Nutanix, Proxmox, Microsoft Hyper V, or public cloud was modeled, costed, and in some cases executed in part. The exit cohort was the smallest in our engagement file, but the most influential on the others.
The Broadcom VMware reset is now the template every prospective buyer should assume the next major acquisition will follow. Perpetual retired, portfolio bundled, renewal repriced, exit path narrowed.
The buyer side preparation for an ownership change at any major vendor on the panel begins the moment the deal is announced, not the moment the rate card changes.
Salesforce announced a price increase in July 2023, the first list move in seven years. The move was followed by a shift in how AI capability was sold, first under Einstein and then under Agentforce.
The Salesforce list pricing update raised list prices on Sales Cloud, Service Cloud, Industry Cloud, Marketing Cloud, and Tableau by an average of nine percent, effective August 2023. The increase landed on new sales and renewals, with timing dependent on existing contract terms.
The buyer side outcome split predictably. Customers in the middle of a multi year deal with a price hold rode the existing rate. Customers up for renewal absorbed the list change, with the realized increase depending on commitment level and term extension.
By 2024 to 2025, Salesforce had repositioned its AI portfolio under Agentforce, with consumption based pricing on agent actions rather than per seat add ons. The economic effect is similar to the AI premium elsewhere on this timeline, but the unit of cost is different.
The new question on a Salesforce renewal is no longer just the seat rate. It is the action volume, the agent template count, and the rate per action, and the answers compound quickly across customer service and revenue use cases.
Beyond the headline list move and Agentforce, Salesforce expanded the platform footprint through Data Cloud, MuleSoft entitlements, and packaging changes on Industries clouds. Each carried its own pricing model, and each tended to enter the contract quietly.
The buyer side discipline was to inventory every entitlement at renewal, identify the lines that had grown without explicit commitment, and either right size them or surface them as negotiation levers against the headline rate.
The reliable move on a Salesforce renewal is to separate the base seat negotiation from the Agentforce consumption term. Bundling the two into one number favors the vendor. Separating them lets the buyer commit to seats while keeping the AI consumption short term and easier to right size.
SAP did not run a single dramatic list move in the period, but two structural changes reshaped the cost path for almost every SAP customer. The first was the support fee handling. The second was the ECC end of mainstream maintenance and the runway toward S/4HANA and RISE.
SAP indexed Standard and Enterprise support increases against an inflation benchmark and committed to keep changes inside that band, with the methodology published on its support services page. In practice the increases were smaller than several headline moves elsewhere on this timeline, often in the low single digits.
The catch is that the base the percentage is applied to is large, and the cumulative effect compounds. A few percent each year on a multi million Euro support base still translates to material absolute growth.
SAP set the end of mainstream maintenance for SAP ECC at the end of 2027, with optional extended maintenance to 2030 at higher fees. The clock effectively forces a decision, either move to S/4HANA on a customer chosen path or move to RISE with SAP, the managed cloud service.
The cost effect is not a single price increase. It is the conversion economics of the migration plus the new commercial model that follows. A RISE deal looks different from an ECC license plus support deal in every dimension, including how the cost is recognized and how scale changes are priced.
SAP also clarified the indirect access measure during the period, moving toward Digital Access at a per document price. The change was framed as simplification, and for some customers it was. For others, particularly those with high transaction volumes from non SAP front ends, the new metric exposed cost that was previously diffuse.
The RISE with SAP commercial model bundles infrastructure, basis services, and SAP S/4HANA into a per FUE subscription. The realized cost depends almost entirely on how the FUE conversion is negotiated, and on how the application volume and tier are mapped.
Buyers who modeled the FUE conversion against their actual usage before signing held the RISE economics close to the on premise baseline. Buyers who accepted the vendor mapping at face value paid materially more for the same workload.
The five vendors covered above account for most of the headline noise, but the timeline runs across the full eleven vendor panel. Four more moved on price in ways that mattered for an enterprise budget.
IBM did not run a single sweeping list move on the order of Oracle Java or Broadcom VMware. The pressure on customers showed up at Enterprise License Agreement renewal, where IBM moved aggressively to bundle Red Hat with the rest of the portfolio after the 2019 acquisition closed.
The realized increase for an IBM customer in the period commonly came from scope expansion, not list rate movement. A buyer who renewed an ELA without right sizing the prior commitment carried unused entitlement at full unit cost into the next term.
ServiceNow does not publish a list rate card the way Microsoft does, so its moves landed inside renewals rather than in press releases. The pattern through 2024 and 2025 was opening quotes in the high teens to mid twenties uplift band, with the realized increase pulled down to a band buyers could absorb.
The Now Assist AI add on followed the same shape as Copilot and Agentforce, an additional priced layer on top of the platform subscription, with a similar effect on the net renewal number.
Adobe added AI tiers to Creative Cloud, Document Cloud, and Experience Cloud during the period, with Firefly generative credits and embedded AI features priced as new SKUs rather than included in the existing seat.
The effective increase for a buyer that attached the new tiers landed in the mid to high teens band on the affected seats, with variability driven by attach ratio and term commitment more than by list movement.
Google Cloud held its standard rates relatively steady on infrastructure during the period, but the contract terms changed. Committed use discounts narrowed, sustained use credits became less generous in practice, and Workspace pricing on the higher tiers rose.
The realized effect on a Google Cloud enterprise customer was less a price hike than a tightening of the negotiated commit structure, which produced the same outcome on a multi year basis.
The same vendor moves did not land the same way across every buyer. Industry, contract maturity, and the renewal calendar each tilted the outcome in directions that are visible in our engagement file.
Financial services buyers carry the largest absolute Oracle, Microsoft, and IBM estates, and the longest contract maturities. The Java metric switch and the VMware reset both hit this sector hardest in dollar terms, simply because the base was largest.
The mitigating factor was contract tenure. Multi year deals with capped uplifts negotiated before the move dampened the blow on the year the metric changed, even as the next renewal carried the new measure.
Manufacturing customers were the most exposed to the SAP runway, the VMware reset, and the Microsoft Copilot attach. The pattern in this sector was a high concentration of long lived enterprise applications on long maintenance tails.
The realized outcome depended on whether the buyer started the S/4HANA or RISE conversation early enough to negotiate from a position rather than a deadline. Late starters often paid in cash conversion as well as in unit rate.
Technology buyers carried the most variable Salesforce, Workday, and AWS footprints. They were among the earliest to attach generative AI tiers and among the most exposed to the consumption based pricing that followed.
The lesson from this sector was that early attach without a usage cap can compound quickly. Buyers who modeled the AI premium against a usage forecast before signing paid materially less per useful output than buyers who did not.
Public sector and regulated buyers tend to renew on long, fixed cycle frameworks. The vendor moves on this timeline mostly arrived at scheduled renewal points rather than mid term, but when they did, the framework constraints often limited the buyer side response.
The reliable lever in this sector was the framework itself. Buyers who built the price hold and benchmarking clauses into the framework rate card carried protection across multiple renewals.
The headline moves on this timeline tell only half the story. The other half sits in what enterprise buyers actually signed, which is rarely the same number as the vendor opening ask. Three composite patterns repeat across the engagement file.
The prepared buyer modeled the new measure before the quote, identified a credible alternative, and started the renewal cycle nine to twelve months ahead. The realized increase across this cohort landed at roughly 40 to 60 percent of the vendor opening ask, with the largest reductions on the steepest moves.
The prepared buyer almost always paid less in absolute terms over a multi year window even when the headline rate looked similar to a less prepared peer. The difference compounded into the next renewal because the base was smaller and the clauses were tighter.
The reactive buyer responded after the first quote, usually two to three months before contract expiry. The realized increase across this cohort landed at 80 to 100 percent of the vendor opening ask, occasionally above it when the deadline ran past expiry.
The reactive pattern accounts for more enterprise spend than most procurement teams recognize. The visible cost was the renewal number. The invisible cost was the leverage forfeited for the next cycle, locked in by the terms signed under deadline pressure.
A subset of buyers used the moves on this timeline as an opportunity to consolidate. Where a vendor reset its model, they accepted the disruption and rebuilt the contract from scratch with new terms, often with a smaller estate and a different mix.
The consolidator pattern produced the largest absolute savings in the period, but only for buyers who had the operational appetite to retire underused products and migrate workloads. Where the operational change held, the financial gain held.
The hike timeline at a glance, 2019 to 2026
| Year | Vendor | Move | Effective range for a typical estate |
|---|---|---|---|
| 2019 | Oracle | Java SE Subscription replaces free commercial updates | New annual line item; first contract often six to seven figures |
| 2022 | Microsoft | First commercial Microsoft 365 list increase in a decade | +9 to 25 percent by SKU; locked in EA, immediate in CSP |
| 2022 | VMware | Subscription bundles begin to replace perpetual offers | Modest base move; the bigger shift came in 2024 |
| 2023 | Salesforce | First list increase in seven years, average nine percent | ~+9 percent on Sales, Service, Industry, Marketing, Tableau |
| 2023 | Oracle | Java moves to employee metric | 2 to 5x effective rise on most estates; far higher on small footprints |
| 2023 | Microsoft | Microsoft 365 Copilot add on at 30 dollars per user per month | ~Doubles the cost of an E3 seat when attached |
| 2024 | Broadcom VMware | Perpetual retired, portfolio consolidated to bundles | +100 percent and up on annualized cost in many cases |
| 2024 to 2025 | Salesforce | Agentforce consumption pricing rolls out | Net new spend line tied to action volume |
| 2025 | Adobe | AI tier add ons across Creative Cloud and Experience | Mid to high teens band, depending on attach |
| 2026 | Microsoft | Targeted commercial step on E3, E5, security add ons | Single digit to low teens by SKU |
| 2026 | Multiple | AI premium becomes the largest source of net new spend | Variable; consumption based across most vendors |
The standard reseller and account team pitch is that each price move is a one off shock, that the right move is to negotiate the headline down at the moment the quote arrives, and that long term contracts with the same vendor are the safest hedge. We disagree. In roughly 200 of our 2024 and 2025 engagements that touched a vendor on this timeline, every major increase was preceded by a public signal months ahead, and the buyers who treated the timeline as the unit of analysis paid far less than those who reacted one quote at a time. The buyer side move is to read the next signal, model the renewal nine to twelve months ahead, and walk in with a credible alternative already costed.
Source: Redress Compliance advisory engagement file, 2019 to 2025.
The vendor sets the date. The signal sets the timeline. The buyer who reads the signal sets the price.
The same three signals that explain the moves since 2019 are likely to explain the next moves. Each one is observable in public reporting and earnings calls, often quarters ahead. The reliable buyer side response is to watch the signals and prepare the renewal calendar around them.
An acquisition closing, a private equity buyout, or a board reshuffle at a software vendor is a near certain signal of a repricing inside twelve months. The pattern has held across Broadcom and VMware, and across several mid market roll ups in security and observability.
The buyer move is to pull the renewal date forward in scope planning, model the conversion economics, and start an alternative evaluation early enough to land before the new owner controls the rate card.
When a vendor changes how it counts users, processors, employees, transactions, or agents, the price is changing even if the headline rate is not. The Oracle Java employee metric, the SAP Digital Access measure, and the Salesforce Agentforce action are all examples on this timeline.
The buyer move is to model the new measure against the actual estate before the quote arrives, and to keep a snapshot of the prior measure as a contractual reference.
Every priced AI add on on this timeline has been announced months before it became material to a renewal. Microsoft 365 Copilot, Salesforce Agentforce, ServiceNow Now Assist, and Adobe Firefly all followed the same pattern.
The buyer move is to treat the AI add on as a separate contract from the base license, with its own term, cap, and exit, and to model attach as a budget line independent of the seat count.
Four clauses together convert an open ended renewal into a bounded one and consistently hold the gap between vendor opening ask and signed cost widest in our engagement file.
The single most consistent factor in the realized outcomes we tracked through this period was the calendar. Buyers who started the renewal cycle nine to twelve months ahead of expiry consistently landed quotes well below the opening ask. Buyers who started after the first quote arrived rarely closed the gap.
The calendar is the cheapest leverage available, because it is free and entirely under the buyer side. Every other lever on this timeline depends on time to model, time to evaluate alternatives, and time to negotiate from a position rather than a deadline.
The same three signals point to a 2027 that looks like a continuation of 2024 to 2026 rather than a break with it. The dominant driver will be the AI premium and the consumption based pricing that follows it. Headline list moves will likely run smaller than the AI attach economics underneath them.
Across the panel, the priced AI add ons that arrived in 2023 to 2025 are now reaching their first renewal anniversaries. Vendors that signed customers on promotional rates will move those rates toward the published list at renewal, with the gap closing rather than the headline moving.
The buyer side exposure in 2027 sits with customers who attached AI add ons broadly in 2024 to 2025 without modeling the renewal terms. The promotional window closes before the productivity gain is fully captured in most estates.
The vendors most likely to run a meaningful list move in 2027 are those that have not yet repriced in the period, or that have a strategic reason to step before competitors. Workday, Cisco, and selected security vendors sit higher in our watch list for 2027 than the headline names.
The vendors least likely to run a fresh list move are those that have already reset their model recently, including Broadcom VMware and Oracle Java. Their 2027 pressure will arrive through scope and attach, not headline rate.
The renewal calendar that worked through 2024 to 2026 will work through 2027. Nine to twelve months ahead, modeled new measures, credible alternatives, and separate AI add on terms remain the load bearing moves.
The only addition is the AI renewal anniversary, which now sits inside many enterprise contracts. Adding the AI attach anniversary to the calendar alongside the base renewal anniversary is the single new discipline 2027 requires.
Oracle raised Java pricing in 2019 and again in 2023, Microsoft 365 in 2022 and again in 2026, Salesforce in 2023, Broadcom VMware in late 2023 into 2024, and SAP applied indexed support increases while running ECC toward end of mainstream maintenance in 2027. The order matches the three signals.
Every major move since 2019 was telegraphed by one of three signals, an ownership change at the vendor, a licensing model or metric change, or the launch of a priced generative AI add on. The signal is almost always public months before the rate change lands. The buyer side move is to prepare the renewal before the quote arrives.
Oracle moved Java to a per employee metric in January 2023, replacing the previous per user and per processor model. Every employee, contractor, and contingent worker of the customer is now in scope regardless of who runs Java. For most estates the change raised the effective cost by a factor of two to five or more.
Microsoft raised commercial Microsoft 365 and Office 365 list prices in March 2022, the first list move since the suites launched. The increases ran from roughly nine to twenty five percent depending on the SKU. A second targeted step landed in early 2026 on selected SKUs including Microsoft 365 E3, E5, and certain security add ons.
Broadcom completed the VMware acquisition in November 2023 and reset the VMware portfolio in December 2023 and January 2024, retiring most perpetual offers and consolidating SKUs into bundles led by VMware Cloud Foundation. Transition quotes commonly landed in the 100 percent plus uplift band on annualized cost, the steepest single move on this timeline.
Three signals predict almost every move on this timeline. An ownership change at the vendor is the most reliable, a licensing model or metric change is the next most reliable, and the launch of a priced generative AI add on is the third. Each signal is observable in public reporting months before the rate change lands.
Yes. Across the vendors on this timeline, enterprise software list prices have risen roughly two to three times faster than general consumer inflation since 2021. The structural drivers, consolidation and AI bundling, sit largely outside the inflation cycle, so the gap is likely to persist even as headline inflation moderates.
Start the renewal calendar nine to twelve months ahead of expiry, model any new licensing measure against the actual estate, and cost a credible alternative for each critical category. Separate any AI add on from the base license with its own term, cap, and exit. Engage independent benchmarking before the first vendor conversation.
Four clauses together hold the gap between vendor opening ask and signed cost widest in our engagement file. A capped uplift tied to a published index, co terminus renewal dates, defined swap and reallocation rights, and a benchmarking reference clause. Buyers who secure all four typically see the smallest gap between list movement and realized cost across the timeline.
The vendor by vendor calendar of moves since 2019, the three signal model that predicted each one, the realized cost bands from our engagement file, and the renewal clause checklist that holds the gap widest.
Used across more than five hundred enterprise engagements. Independent. Buyer side. Built for procurement, sourcing, and finance leaders running the next renewal cycle.
The vendor sets the date. The signal sets the timeline. The buyer who reads the signal sets the price.
Tell us which vendor and renewal anniversary is next. We will model the move against the timeline and the engagement file and reply within one business day.
Contact Us →A short read on the vendor signals that move enterprise software prices. Delivered weekly to procurement, sourcing, and finance leaders.
