Enterprise software prices rose faster than inflation again in 2026. This index tracks the move across eleven major vendors, separates headline list increases from what buyers actually pay, and isolates the part that is purely the AI premium.
Enterprise software prices rose faster than inflation again in 2026. This index reads the move across eleven major vendors, separates the headline list increase from what buyers actually sign, and isolates the share that is purely the AI premium.
About this index
The Redress Compliance Enterprise Software Price Increase Index is a directional benchmark, 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 range rather than a guarantee.
The blended answer is roughly 8 to 12 percent for the year, and about 48 percent over five years. Our composite index reached 148 in 2026, measured against a base of 100 in 2021. That is a compounding rate well above the general price level.
The headline matters less than the spread beneath it. A single blended number flattens a market that ranged from low single digit uplifts to outright model resets. Read the index as the weighted center of gravity, not as the figure any one buyer should expect on a given renewal.
The index weights each vendor by its share of enterprise software spend in a typical large estate, then blends the published list move with the realized move our panel actually paid. It is a cost of ownership signal, not a vendor revenue figure.
This matters because revenue growth and buyer cost can diverge. A vendor can grow revenue through new logos while a given customer holds flat, or hold revenue flat while repricing an existing base hard. The index follows the existing base, which is where most enterprise budget sits.
List price is the sticker. Realized cost is the signature. The two move together over long periods, but in any single renewal they can sit 10 to 20 points apart depending on how the buyer runs the cycle.
Most published price trackers report only the list move, because that is the public number. Our index reports both, because the buyer only ever pays the second one. The wedge between them is the entire value of a disciplined renewal.
Two forces stacked in 2026. The subscription resets that began in 2024 reached more renewal anniversaries, and the AI add on layer moved from pilot pricing to full list pricing. Both pushed the realized number up at once.
Gartner has framed the macro backdrop plainly. It forecasts worldwide IT spending above six trillion dollars in 2026, and has noted that a large share of software spend growth now reflects price increases rather than new deployment. Our index is the buyer side view of that same shift.
The vendor moves do not happen in a vacuum. They sit inside a software market growing faster than almost any other line in the corporate budget, and growing for reasons that favor the seller.
Gartner expects worldwide IT spending to pass six trillion dollars in 2026, with software the fastest growing major segment. A meaningful share of that growth is price, not new deployment. It is the same software, repriced.
Analysts have attributed a large part of software spend growth to inflation in the price of existing products rather than expansion into new ones. When most of the growth is repricing, the buyer funds the increase out of the same budget, for the same capability.
Surveyed technology leaders have reported expecting cost increases near nine percent for IT products and services. That expectation becomes self fulfilling. Budgets are set to absorb it, and vendors price into the room they are given.
The fastest growing slice is generative AI. Spending on AI models and AI enabled software is rising at a pace no other category matches, and it is additive. It rarely replaces an existing line. It stacks on top.
That is the macro version of the AI premium we isolate later. At the market level it shows up as a category growing many times faster than the software market around it, pulling the blended index upward.
A blended index understates the pain for a specific buyer for two reasons. First, no buyer holds the market basket. They hold a concentrated set of contracts, often weighted toward the vendors that moved most.
Second, increases compound. An estate weighted toward a vendor that reset its model, or toward AI heavy seats, runs well ahead of the 148 reading. The index is the market average. The budget is a specific, and usually less favorable, slice of it.
Because the main drivers sit outside the inflation cycle. Wage and input costs explain a few points. The rest is structural: who owns the software, how it is packaged, and what new layer is being attached on top.
Fewer independent vendors means fewer credible alternatives at the table. Each large acquisition removes a competitive option and hands the acquirer more pricing power over the installed base.
The pattern is consistent. After a major platform is acquired, the new owner narrows the product line, raises prices on what remains, and reduces the discount latitude field reps can offer. The buyer feels it at the first renewal after close.
Converting a perpetual license to a subscription is a price increase wearing a modernization label. A fully amortized perpetual asset carried only a maintenance fee. The subscription replaces that small fee with a full recurring license cost.
For many estates this is the single largest line item change in a decade. The cash impact is immediate, and there is rarely a path back to the old model once the perpetual rights lapse.
When a vendor changes the unit it counts, the price can rise sharply without the rate card moving at all. A shift from named users to total employees, or from cores to bundles, resets the entire calculation.
Oracle's move to a per employee metric for Java SE Universal Subscription is the textbook case. The per unit price looked modest. The unit count expanded to the whole organization, and the bill rose with it.
Generative AI add ons are the newest and fastest growing driver. They are sold on top of the base license, not inside it, so they expand spend without touching the existing rate card.
This is why the AI premium deserves its own section below. It is not a discount problem or a renewal problem. It is a new layer of spend that did not exist three years ago and is now compounding faster than anything else in the stack.
The spread is the story. One vendor reset its entire model. A cluster sat in the high teens. The largest names by spend mostly landed in single digits to low teens, and a quiet group held nearly flat.
Broadcom ended new perpetual VMware license sales in early 2024 and collapsed thousands of products into a handful of subscription bundles. For customers moving off perpetual licenses, the annualized cost change ran well above 100 percent, and public cases reached far higher.
This is not an annual uplift. It is a one time reset of the cost base that then carries forward. We treat it as the off scale anchor of the index, not as a comparable percentage.
Oracle continued to lift effective cost through metric and audit pressure, landing many estates in the 10 to 20 percent range on a blended basis. ServiceNow renewals frequently opened at double digit uplifts, with AI add ons pushing the effective figure higher.
Adobe restructured Creative Cloud into tiers, with the full AI capable tier carrying effective increases that reached the mid twenties for users who needed the complete feature set. The common thread is packaging, not a simple rate card move.
Microsoft confirmed its first broad commercial increase since 2022, with Microsoft 365 business plans stepping up from 1 July 2026. The base move was modest. The Copilot add on did the heavy lifting on total cost.
Salesforce raised list prices about 9 percent in its first increase in seven years, then layered AI on top. SAP applied an on premise support fee increase capped near 5 percent and steered customers toward RISE with SAP. Workday held to mid single digit to low double digit renewal uplifts.
IBM increases clustered in the low to mid single digits on most renewals, with sharper moves where a metric or product line changed. Cisco applied a portfolio wide uplift in the low single digits across hardware and technical services.
The hyperscalers are commitment driven rather than list driven. Effective cost moved with consumption and committed spend tiers, so a buyer's outcome depended more on the deal structure than on any published rate.
2026 movement at a glance, by vendor
| Vendor | 2026 direction | Realized band | Dominant driver |
|---|---|---|---|
| Broadcom / VMware | Model reset | 100%+ on transition | Perpetual to subscription |
| Adobe | Up sharply on AI tiers | Up to ~27% | Tier restructure plus AI |
| ServiceNow | Up | 10 to 20% | Renewal uplift plus AI |
| Oracle | Up | 10 to 20% | Metric and audit pressure |
| Microsoft | Up | 8 to 15% | Base step plus Copilot |
| Salesforce | Up | 9% list, plus AI | First list move in 7 years |
| SAP | Up | 5 to 10% | Support cap plus RISE |
| Workday | Up | 5 to 10% | Renewal uplift |
| IBM | Up modestly | 4 to 8% | Renewal and metric changes |
| Hyperscalers | Commitment driven | Flat to ~10% | Consumption and commit tiers |
| Cisco | Up modestly | 3 to 5% | Portfolio wide uplift |
Four or five vendors account for the bulk of software spend in most large estates. Their individual decisions shape the index more than any average. Here is what each did, and what it means for the bill.
Oracle raised effective cost less through the rate card and more through the unit it counts and the audits it runs. The per employee Java metric is the clearest example, expanding the count to the whole organization regardless of who uses the software.
The buyer side reality is that an Oracle increase often arrives as an audit finding rather than a renewal letter. The defense is an estate baseline built before the conversation, not after the finding.
Microsoft confirmed its first broad commercial price step since 2022, with business plans rising from 1 July 2026. Business Basic and Business Standard both move up, while the premium tier holds. The base move is real but modest.
The larger number is Copilot. At thirty dollars per user on an annual commitment, the AI add on can rival the cost of the underlying seat. For Microsoft, the 2026 increase is mostly an attach story, not a rate card story.
SAP applied an on premise support fee increase tied to local inflation with a cap near five percent, a step up from the prior cap. At the same time it tightened the runway on older ECC releases.
Standard support for ECC steps down after 2027, with extended maintenance available through 2030 at a premium. The pricing is set to make the move to RISE with SAP look like the rational choice. That is an increase, dressed as a migration.
Salesforce raised list prices about nine percent in 2023, its first increase in seven years, lifting the core editions across the board. It then layered Agentforce on top as a consumption priced AI tier.
For a Salesforce heavy estate, the 2026 cost picture is the 2023 list move now fully flowing through renewals, plus whatever AI consumption the business has switched on. The base and the premium arrive together.
Broadcom reshaped VMware faster and harder than any vendor in the index. It cut a catalog of thousands of products to a handful of bundles, ended new perpetual sales, and discontinued lower tiers across 2024 and 2025.
The result for customers on perpetual licenses was a step change, not an uplift. Annualized costs commonly more than doubled, and a number of large accounts faced far steeper proposals. The migration conversations this triggered are still running across the market.
ServiceNow combines a steady renewal uplift with aggressive AI add on pricing. Renewals frequently open in the double digits, and the AI layer can add a further large premium on top of the platform seat.
The control point is rightsizing. ServiceNow estates drift toward higher tiers and unused entitlements, and the renewal is the moment to reset both before the uplift is applied to an inflated base.
It is the difference between the quote and the signature, and it is large. In our engagement data the realized increase after a structured negotiation usually lands at 40 to 60 percent of the vendor's opening ask.
The gap is not luck. It is the product of preparation, timing, and leverage. Buyers who treat the opening quote as a fact pay close to it. Buyers who treat it as a position pay far less.
Renewal quotes are anchored high on purpose. The first number sets the reference point, and an unprepared buyer negotiates down from it rather than up from a benchmarked target.
The fix is to arrive with your own anchor. A benchmarked target, supported by comparable deals, resets the conversation before it starts and removes the vendor's first mover advantage.
Four things move the realized number more than anything else. None of them are clever tactics at the table. They are decisions made months before the table.
More than the headline suggests, because it changes the shape of the cost, not just the size. A perpetual license is a capital asset that depreciates. A subscription is an operating cost that never ends.
Under the perpetual model, a buyer who had finished paying for the license carried only an annual maintenance fee, often around a fifth of the license price. Under subscription, that small fee is replaced by a full recurring license charge.
That is the step that catches finance teams. The line that used to be a modest maintenance accrual becomes one of the larger recurring software costs in the budget, and it does so in a single renewal.
Once perpetual rights lapse or support ends, there is rarely a route back to the old model. The buyer cannot return to maintenance only economics, which removes the fallback that used to cap cost in lean years.
This is why the conversion is so valuable to vendors and so costly to buyers. It is not a one year increase. It resets the floor for every year that follows.
Where a conversion is unavoidable, the move is to negotiate the entry price, the cap on future uplifts, and the exit terms as a single package, while the vendor still wants the signature. Leverage is highest at conversion and falls afterward.
Enough that it now leads. The AI premium is the extra cost of a generative add on sold on top of the base license, and it is the largest single source of net new software spend in the index.
Because it sits outside the base rate card, the AI premium is easy to miss in a renewal model that only tracks the existing line items. It arrives as a separate add on, and it compounds.
Microsoft lists Microsoft 365 Copilot as a 30 dollar per user add on on an annual commitment. On a typical knowledge worker base seat, that add on can roughly double the per user cost.
The decision is not whether Copilot has value. It is whether the attach rate matches actual usage. Paying for every seat when a fraction use the feature is the most common way the premium runs ahead of the benefit.
Salesforce prices its agent layer through Agentforce on a consumption basis, with a flexible credit model layered on the platform license. Consumption pricing can be efficient, but it shifts the budgeting problem from a fixed seat to a variable meter.
That variability is the risk. A pilot that looks cheap at low volume can scale into a material line item once agents move into production traffic, and the meter does not pause for a budget cycle.
The common error is treating AI add ons as an all or nothing seat decision. The premium is controlled by attach discipline, usage telemetry, and a willingness to right size the add on at each renewal, not by refusing the technology.
The standard advice is to consolidate spend with fewer vendors to earn volume discounts and simpler renewals. We disagree, at least as a default. In a large share of the estates we have benchmarked, deep consolidation removed the buyer's last credible alternative and handed the surviving vendor the pricing power that produced the next increase. The discount won at signing was smaller than the leverage lost over the term. The buyer side move is to keep at least one funded alternative alive in every critical category, even at a small efficiency cost, because the option to walk is worth more than the volume tier. Consolidate for operational reasons, not as a price strategy.
Source: Redress Compliance advisory engagement file, 2024 to 2025.
A price increase is a number the vendor proposes. The bill is a number the buyer agrees to. The whole craft of the renewal lives in the distance between the two.
The realized increase is decided by the contract more than the conversation. A renewal with the right clauses bounds the next increase before it is proposed. A renewal without them leaves the buyer exposed every cycle.
The single most valuable clause is a cap on the annual increase, expressed as a fixed percentage rather than a reference to a moving list price. A fixed cap converts an open ended renewal into a known one and removes the vendor's largest lever.
The detail matters. A cap tied to a published index can still climb fast, so the strongest version is a flat numeric ceiling that holds across the full term.
Aligning all renewal dates to a single anniversary prevents the staggered increases that let a vendor reprice one product at a time. A price hold across the term locks the rate for the period rather than only the first year.
Buyers who leave dates scattered hand the vendor a series of small, separate negotiations. Each one is easier for the vendor to win than a single consolidated event the buyer controls.
The right to swap unused licenses for ones the business actually needs, or to reallocate across the estate, protects against paying for shelfware while buying more elsewhere. It is one of the most underused clauses in enterprise contracts.
The newest essential clause treats the AI add on as its own agreement, with its own term, cap, and exit. Bundling AI into the base renewal hides the premium and removes the ability to right size it at the next cycle.
Three anonymized patterns from the engagement file show how the same index plays out differently depending on preparation. The figures are bands, not exact outcomes, and details are generalized to protect confidentiality.
A large financial services buyer received a subscription transition quote that more than doubled its annualized VMware cost. The first instinct was to absorb it, because the platform was deemed too embedded to move.
A funded migration assessment, run in parallel with the negotiation, changed the posture. The credible alternative, though never executed, brought the realized increase down materially from the opening ask and won a multi year cap.
A manufacturer discovered that a per employee software metric had quietly expanded its liability far beyond actual usage. The vendor's opening position counted the entire workforce.
A clean baseline of who actually used the software, built before the response, narrowed the defensible count. The settlement landed well below the opening exposure, and the estate moved partly to a free alternative to cap future drift.
A retailer was offered an AI add on across its entire knowledge worker base. At full attach, the premium rivaled the cost of the underlying seats themselves.
Usage telemetry from a measured pilot showed real demand in a fraction of roles. Attaching the add on to that fraction, with a term and a cap, captured the value while holding the premium to a small share of the all in quote.
By building the increase into the plan as a scenario band, not a single line, and by funding the work that holds the increase down. A flat software budget in this market is a budget that will be missed.
The index suggests a planning band of roughly 8 to 12 percent on the existing software base, before any new projects. Estates weighted toward the vendors that reset their models should plan toward the top of that band or above.
A single point estimate invites a miss in either direction. A band lets finance hold a contingency for the contracts most likely to reprice and release it when a renewal closes below the ceiling.
The AI premium behaves differently from the base, so it should be budgeted separately. It is newer, less predictable, and tied to attach and consumption rather than seat counts that finance already knows.
Treating it as a distinct line makes the premium visible and reviewable. Folding it into the base hides the fastest growing cost in the stack inside a number that looks familiar.
The cheapest way to lower the realized increase is to fund the preparation that produces it. A benchmark, an early start, and a costed alternative repay their cost many times over across a large renewal.
Some renewals are quiet and some are not, and the difference is usually visible months ahead. A handful of signals reliably flag a renewal that will open high.
The clearest signal is a vendor that has just been acquired or has just shifted strategy. New owners reprice the installed base, so a platform that changed hands in the last two years should be treated as a high risk renewal by default.
The same applies when a vendor announces a move to subscription only or a new packaging model. The announcement is the warning. The renewal is where it lands.
When a vendor changes the unit it counts or collapses products into bundles, the next renewal carries the increase even if the rate card looks stable. Watch for any change to how the contract is measured.
These changes are often buried in updated ordering documents rather than announced. Reading the new terms before the renewal, not during it, is the only reliable defense.
An estate that has grown its usage faster than its entitlement is carrying a hidden increase. The renewal, or the audit that precedes it, is where the gap is priced. Drift is a self inflicted signal, and the most controllable one.
The most valuable thing a buyer brings to a renewal is a real option to leave. It does not have to be exercised. It has to be credible, costed, and visible at the table.
A vague threat to switch carries no weight. A costed migration plan, with a timeline and a number, changes the conversation because the vendor can no longer assume the buyer is captive.
The work of costing the alternative is also the work of understanding your own estate. That understanding is what produces the benchmark and the negotiation position, whether or not you ever move.
A limited proof of concept on the alternative is often enough. It demonstrates feasibility without the cost and disruption of a full switch, and it gives the buyer evidence rather than assertion.
For many categories the alternative is an open source or lower cost option the engineering team can stand up quickly. The existence of a working proof is worth more at renewal than its eventual adoption.
An alternative the vendor does not know about provides no leverage. The point is not to bluff. It is to make the option visible and credible so the renewal is priced against a buyer who can walk, not one who cannot.
Earlier than most buyers think. The best outcomes in our panel started 9 to 12 months before expiry. The worst started inside 60 days, when the only remaining options were to accept or to lapse.
Lead time is leverage because it creates the space to build a benchmark, cost an alternative, and let the vendor know both exist. A buyer with a year can walk. A buyer with a month cannot, and the price reflects it.
Vendors understand this calendar better than buyers do. The late stage discount that appears near quarter end is real, but it is offered to a buyer who has already lost the time to use it fully.
The fix is a standing renewal calendar that flags every contract a year ahead. Built once and maintained, it converts every renewal from a surprise into a planned event with time to prepare.
This single discipline does more for realized cost than any tactic at the table. The buyers who hold the gap widest are simply the buyers who started first.
The headline index is an average, and averages hide where the pressure actually falls. Sector and size both shape how hard a given increase lands and how much room a buyer has to push back.
Regulated industries felt the subscription and AI moves most acutely, because compliance requirements limit how freely they can switch platforms or defer upgrades. That reduced their walk away leverage at exactly the moment vendors repriced.
Technology and digital native firms fared better. They carry the engineering capacity to adopt open alternatives and to migrate, which keeps a real option on the table and holds realized increases lower.
The largest enterprises secured the deepest discounts but absorbed the biggest absolute increases, because their spend base is large and vendors defend those accounts hardest. Discount depth and increase exposure rose together.
Mid market buyers often saw the steepest percentage moves. They lack the spend to command attention, yet they sit above the threshold where standard list pricing applies without negotiation. That band is where unmanaged increases concentrate.
A benchmark is only useful if its limits are stated. This index is a directional read on the cost of owning the same software over time. Several things sit outside it on purpose.
The index tracks recurring license and support cost, not the one time cost of implementation, migration, or professional services. Those are real and often large, but they are project costs, not the price of the software itself.
We exclude them so the index stays comparable year to year. Mixing project spikes into a price signal would make the trend impossible to read.
The index reflects the vendor's own pricing path. It does not assume a buyer takes third party support, an open source alternative, or a re platforming, all of which can change the picture materially.
Those options are exactly what the buyer side response is about. The index measures the default path so the value of leaving it is visible by comparison.
Pricing varies by region and currency, and a global estate will see effects the index does not localize. A buyer should read the index as a direction and then localize it to their own currency, region, and product mix.
This is why we publish bands rather than precise figures. A single decimal would imply a precision that a cross vendor, cross region benchmark cannot honestly claim.
Price control fails most often as an ownership problem, not a negotiation problem. When no one owns the renewal calendar, every renewal arrives as a surprise and is handled in a hurry, which is exactly when buyers overpay.
The renewal calendar needs one accountable owner, usually in procurement or a software asset management function, who is measured on realized cost against benchmark rather than on closing deals quickly.
That measure matters. An owner rewarded for speed signs early and high. An owner rewarded for realized cost starts early and pushes, which is the behavior that holds the gap open.
Finance sets the budget band and the contingency. IT confirms what is actually used and whether an alternative is feasible. Both need to be in the conversation 9 to 12 months out, not consulted at signature.
The common failure is a renewal that reaches finance and IT only when the vendor needs a signature. By then the timeline itself has become the vendor's strongest lever.
An independent advisor adds the benchmark and the market view that no internal team sees across enough deals to hold. The highest return point to engage is before the first response, when the position is still open.
Up again, but with a different mix. We expect another high single digit to low double digit blended increase, with the AI premium as the dominant driver rather than base inflation or new list moves.
Vendors that have already reset their models will lean on attach and consumption rather than headline rate changes. The list cards may look calm while the realized bill keeps climbing through add ons and usage.
That makes 2027 harder to read from public price trackers alone. The increase will hide inside consumption meters and AI attach, which is precisely where a list only view goes blind.
The buyers most exposed are those who signed AI add ons in early promotional windows. As introductory pricing expires, the renewal of the AI layer itself becomes the next repricing event, on top of the base renewal.
The defensive move is to treat every AI add on as a contract with its own term, cap, and exit, not as a feature toggle. The buyers who do that will keep the 2027 premium bounded.
A benchmark is only as good as the comparables behind it. A number with no basis is just an opinion, and a vendor will treat it as one. A defensible benchmark rests on comparable deals, recent timing, and like for like scope.
The strongest benchmark compares your deal to others of similar size, industry, and product mix, not to a published average. A figure drawn from genuinely comparable buyers is hard for a vendor to dismiss, because it reflects what its own peers actually paid.
Generic averages invite the response that your situation is different. Comparable references remove that argument, because the comparison is already adjusted for the things that make deals differ.
Pricing moves quickly in this market, so a benchmark from two years ago understates today's reality and weakens your position. The reference set has to be recent enough to reflect the current rate environment.
This is one reason an internal benchmark drifts out of date. A team sees its own renewals only every few years. A benchmark needs the flow of deals that a continuous, cross client view provides.
A credible benchmark normalizes for scope. It compares the same metric, the same term length, and the same support level, because a lower headline rate on a different scope is not a real comparison.
Getting scope right is unglamorous and decisive. Most benchmark disputes are really scope disputes, and the buyer who has normalized carefully tends to win them.
Few buyers can build a benchmark like this alone, because no single buyer sees enough comparable, recent deals to populate it. This is the core of what an independent benchmarking practice provides.
The value is not a single number. It is the continuously refreshed set of comparables behind it, and the scope normalization that makes the number stand up when a vendor pushes back.
Not all levers are equal. A few move the number materially, and the rest are noise. Knowing which is which keeps a negotiation focused on what actually changes the bill.
A longer term traded for a capped rate is often the highest value lever, because it locks the floor while the vendor still wants the commitment. Timed to the vendor's quarter or year end, the same commitment buys more.
The caution is that term length without a cap simply defers the increase. The lever is the combination, the term and the cap together, not the term alone.
Committing more volume can earn a better unit rate, but it cuts both ways. Volume that removes your last alternative weakens the next renewal even as it improves this one.
Use volume as a lever where it does not cost you leverage. Avoid it where the discount won is smaller than the option lost, which is the contrarian point made earlier in this report.
The most overlooked lever is buying less. Right sizing the estate before the renewal removes shelfware from the base the increase is applied to, and a smaller base compounds into a smaller bill every year after.
This lever is entirely in the buyer's control and needs no vendor agreement. It is the cheapest increase to avoid, because it is the spend you were never using.
On a blended basis enterprise software prices rose roughly 8 to 12 percent in 2026, with the composite index reaching 148 against a 2021 base of 100. That headline hides a wide spread. Some vendors held to low single digit uplifts while others reset their pricing models. The blended figure is the weighted middle, not what one buyer should expect.
Broadcom VMware is the clear outlier. Transition quotes from perpetual licenses to the new subscription bundles commonly landed well above 100 percent, and a small number of public cases ran far higher. No other vendor we track came close. The next tier, with effective increases in the high teens to mid twenties, included Adobe AI tiers and ServiceNow renewals.
Yes. Across the vendors in this index, software list prices rose about two to three times faster than general consumer inflation in 2026. Gartner has noted that a large share of software spend growth now reflects price increases rather than new deployment. The structural drivers, consolidation and AI bundling, sit largely outside the inflation cycle.
The AI premium is the extra cost of a generative add on layered on top of a base license. For Microsoft 365 it is the 30 dollar Copilot add on, which can roughly double the cost of a knowledge worker seat. For Salesforce it appears as Agentforce consumption. The premium is the largest source of net new software spend.
List increases are an opening position, not a final bill. The realized increase after a structured negotiation typically lands at roughly 40 to 60 percent of the vendor's opening ask in our engagement data. The size of the gap depends almost entirely on buyer preparation, timing, and whether a credible alternative exists.
In effect, yes. Converting a fully amortized perpetual license to a subscription resets the annual cost from a small maintenance fee to a full recurring license fee. The headline may be framed as modernization, but the cash impact on the buyer is a step change in annual cost. Broadcom VMware is the clearest 2026 example.
We expect another high single digit to low double digit blended increase in 2027, with the AI premium as the dominant driver rather than base inflation. Vendors that have already reset their models will lean on attach and consumption rather than headline list moves. The buyers most exposed are those renewing AI add ons signed in early promotional windows.
Partly. A multi year deal with a capped annual uplift is the most reliable protection against repricing, but only if the cap, the renewal terms, and the swap rights are negotiated up front. A multi year term with an uncapped renewal at the end simply defers the increase. The protection is in the clauses, not the term length.
The price hold or capped uplift clause matters most, followed by co terminus renewal dates, defined swap and reallocation rights, and a benchmarking reference. Together these convert an open ended renewal into a bounded one. Buyers who secure all four typically see the smallest gap between list movement and realized cost.
Start with a benchmarked target and a credible alternative before you respond. The opening quote is almost always above what comparable buyers pay, and an early, evidence based counter resets the conversation. Engaging an independent advisor before the first response, rather than after the deal stalls, is the single highest return move.
The vendor by vendor movement bands, the list versus realized model, the AI premium calculator inputs, and the renewal clause checklist that holds the gap widest.
Used across more than five hundred enterprise engagements. Independent. Buyer side. Built for procurement and finance leaders running the next renewal cycle.
The vendors set the list. The market sets the floor. The buyer who knows both, and starts early, pays the floor.
