The Three-Year Deal Is Dying: Why B2B Buyers Tripled Sub-One-Year Contracts in 2026 — and the Renewal Math Quietly Rewriting NRR

"We'd love to do three years. We just can't sign for three years right now."

That's roughly the sentence — almost verbatim, every time — that B2B procurement leaders have been saying on closing calls since the back half of 2025. The vendor pitches a multi-year deal with a 22% prepay discount. The slide deck makes the math look obvious. The CFO on the buyer side smiles politely, asks a single question about AI roadmap risk, and then proposes a 12-month term with a renewal option. Sometimes a 6-month pilot. Increasingly, month-to-month with a 90-day notice clause.

The vendor's CRO leaves the call thinking it was a one-off. It wasn't. It was the shape of the market.

For Chief Revenue Officers, Chief Financial Officers, Heads of Deal Desk, Pricing Leaders, Revenue Operations Leaders, and B2B Founders building 2026 financial plans against an installed base that suddenly will not commit past twelve months, the contract-term compression of 2025–2026 is the most underreported structural shift in SaaS economics. It is not a discounting story. It is not a procurement-fad story. It is a permanent rewiring of how enterprise software gets bought — and the companies still modeling growth on three-year deals are about to discover that their forward ARR is sitting on a foundation that buyers have already quietly walked away from.

The Numbers Tell a Story Most Boards Haven't Caught Up To

The cleanest data point on the shift comes from ICONIQ's State of Go-to-Market 2026, which surveyed more than 150 B2B GTM executives in January 2026. Inside the report sits a single chart that, once you see it, becomes hard to unsee.

In 2023, 4% of new logo contracts signed by the surveyed B2B SaaS companies were under 12 months. By 2026, that number was 13% — a more-than-tripling in 30 months. Over the same window, three-year contracts dropped from 28% to 23% of new deals. The annual deal — the workhorse 12-month commitment that has been the median SaaS contract for a decade — is being squeezed from both directions, but the squeeze is asymmetric. Buyers are not migrating toward longer terms. They are migrating toward shorter ones.

Layer in the broader market context and the picture sharpens. Gartner's enterprise SaaS contract sample now splits roughly 42% monthly, 45% annual, and 13% multi-year or usage-based — the closest the market has come to a true split decision on commitment length since the cloud transition began. Average B2B sales cycles compressed from 25 weeks in H1 2025 to 19 weeks in H2 2025, but not because buyers got more confident. They got faster precisely because they were committing to less. Shorter cycles. Shorter terms. Smaller downside.

The pattern is consistent across revenue bands. Buyers in the $5M ARR cohort and buyers in the $500M ARR cohort are asking for the same thing in slightly different language: flexibility now, optionality later, no multi-year handcuffs until you prove the AI roadmap isn't going to make this product obsolete in 14 months.

Why Buyers Suddenly Refuse to Sign for Three Years

The instinct inside vendor finance teams is to read the shift as a temporary procurement reflex — a hangover from 2023 budget tightening that should resolve as confidence returns. That read is wrong. The drivers are structural, and three of them are reinforcing each other.

The first is AI replacement risk. Across nearly every horizontal SaaS category — analytics, marketing automation, BI dashboards, knowledge management, contract review, customer support — buyers can now see a credible scenario where a vertical AI agent or an in-house Copilot eats 40% to 70% of the workflow inside 18 months. They don't necessarily believe it will happen. They believe it might. And once "might" enters the room, the math on a 36-month commitment changes immediately. The CFO is no longer optimizing for the discount. The CFO is optimizing for the option value of being able to walk.

The second is pricing-model volatility. The shift from per-seat to usage-based to outcome-based pricing has made the cost of a multi-year deal harder to model, not easier. 78% of IT leaders now report receiving unexpected charges under hybrid pricing structures that combine seats, consumption credits, and AI surcharges. When the bill itself is uncertain, the commitment length collapses. A three-year deal on an uncertain unit price is just three years of uncertain bills. Buyers would rather renegotiate the unit price every twelve months than lock it in once and ride it.

The third is the buying committee structure itself. The median B2B buying group for deals over $50K now averages 11.2 stakeholders, up from 9.7 in 2024, with 90% of successful deals requiring multi-threading across 5 to 25 buying committee members. Getting consensus on a 12-month commitment is hard. Getting consensus on a 36-month commitment is almost impossible — every additional stakeholder is another veto point, and every additional year of commitment is more political risk for each stakeholder to absorb. Buying committees do not optimize for vendor ARR. They optimize for minimum group regret. Short contracts produce less regret in either direction.

Stack the three together and the shift stops looking like a fad. It looks like the new normal.

The Discount Ladder Has Stopped Working

For the better part of a decade, the multi-year deal was held together by a single instrument: the prepay discount. Sign for two years, get 15% off. Sign for three years, get 22% off. The median enterprise multi-year discount in 2024 sat at 18%, with the most aggressive vendors — Atlassian, Okta, and a handful of others — pushing up to 25% for three-year rate locks. The instrument worked because buyers traded a future option for a present-day cash saving, and the saving was meaningful enough to override the optionality.

That instrument is bending. Not because the discounts got smaller — they didn't — but because the buyer-side value of optionality went up. When the perceived risk of being stuck on the wrong vendor in month 22 is high enough, no two-digit discount makes the option worth giving up. The discount ladder is being repriced by the option premium, and most vendor finance teams have not adjusted their target discount thresholds to reflect that.

Buyers have responded by inventing a new layer of contract instruments that vendors are still learning to negotiate against. The three most common, all of which appeared in less than 5% of enterprise SaaS contracts in 2023 and now appear in well over a third:

• Mid-term re-rate clauses — the right for the buyer to renegotiate per-seat or per-unit pricing at month 6 or month 12 if adoption falls below a stated threshold.
• Future service credits — a portion of multi-year prepayment held in escrow, redeemable against features still on the vendor roadmap.
• Off-ramp triggers — termination rights activated by specific events (acquisition, regulatory change, AI-driven category disruption) that effectively convert a three-year deal into a series of one-year deals with vendor downside risk attached.

The pattern is consistent: buyers are buying optionality even inside the multi-year deals they still sign. The contract length looks the same on the cover page. The actual commitment is shorter than the cover page suggests.

What This Does to Vendor Economics — and Why Most Finance Models Haven't Re-Priced It Yet

The financial knock-on effects of contract-term compression are larger than they appear, and most CFO models built before mid-2025 do not yet reflect them. Three are worth naming directly.

Forward ARR coverage gets weaker. Multi-year contracts are the load-bearing wall of "contracted ARR" — the line on the board deck that boards rely on to forecast 12-month and 24-month revenue floors. As the mix shifts from 28% three-year to 23% three-year and from 4% sub-one-year to 13% sub-one-year, the contracted floor shortens. The same total ARR number now extends 18 months into the future instead of 30 months. That is a real change to valuation math, not a presentation tweak. Public-market analysts have already started discounting forward revenue accordingly.

CAC payback period elongates relative to commitment. The median B2B SaaS CAC payback period has drifted from roughly 11 months in 2021 to 18 months in early 2026. When the median new logo contract was 24 months, an 18-month payback meant the customer paid the vendor back inside the original commitment, with months 19-24 sitting as guaranteed profit. When 13% of new logos commit to less than 12 months, a 17-18 month payback period means the vendor is structurally paying to acquire customers who churn before they break even. That is not a temporary efficiency problem. It is a unit-economics problem hiding inside a length-of-contract problem.

Net revenue retention takes on a new shape. Median NRR across B2B SaaS has compressed from a 2022 peak near 125% to roughly 107% by late 2024 — and as contract terms shorten, the NRR clock moves faster. Customers reach renewal sooner, more often, and with less locked-in lift from expansion built into the original term. Shorter contracts make NRR easier to measure and harder to manufacture. Vendors that were quietly leaning on contractual expansion clauses to flatter their retention numbers are about to find that the contractual expansion no longer covers the gap.

Put bluntly: a vendor with 23% three-year, 64% one-year, and 13% sub-one-year contracts is not running the same business as a vendor with 28% three-year, 68% one-year, and 4% sub-one-year contracts. The first business has more renewal events per dollar of ARR, more touchpoints with a buying committee that has shrunk the time horizon, and far less ability to absorb a bad quarter without it showing up in NRR. The investor coverage models for SaaS have not yet fully repriced this, but the operators on the ground feel it every quarter.

The Renewal Engine Is Now the Growth Engine

The strategic conclusion that follows from the data is the one most vendor leadership teams are slowest to internalize: in a market where contract terms are compressing, the renewal motion stops being a customer success function and starts being the dominant growth function in the company. If the average new logo commits to twelve months instead of twenty-four, the vendor doesn't have two years to land, expand, and prove value before the buyer can walk. They have eleven. After month eleven, the deal is back on the table — and on every renewal call, the buyer has the same AI replacement risk question they had on the original close, except now they also have eleven months of usage data, expense reports, and internal politics arguing against the renewal.

The companies pulling ahead in this environment are doing four things that vendors still running the 2022 playbook are not.

They are pricing optionality, not duration. Instead of selling a three-year deal at a 22% discount, they are selling a one-year deal with a price-locked renewal option, sometimes attached to a usage threshold. The buyer gets the optionality they want. The vendor protects the unit economics. The discount paid is for predictability, not for duration. This is a deliberate inversion of the historical model, and the early signals on win rates suggest it materially closes the gap with buyer expectations.

They are designing the renewal as a quoted-product moment, not an admin moment. Renewals are now run by named owners with quotas — sometimes a Renewal Manager, sometimes a Strategic Account Director — who own the renewal conversation as a deal, not a default. The companies doing this are seeing renewal close rates 8 to 14 points higher than companies still running renewals as a back-office task routed through CS.

They are publishing usage and outcome data inside the contract window. Buyers who refuse three-year commitments will often accept them if the vendor agrees to publish quarterly outcome data — adoption, time-to-value, dollars saved or generated — that supports the value case. The data publication transfers the burden of proof to the vendor and reduces the buyer's sense that they are locked in blindly. The vendors that have built this into the contract are seeing multi-year mix stabilize, even as the broader market drifts shorter.

They are repricing the deal desk function as a strategic pricing function. The companies winning the compression are not running the deal desk as an approvals workflow. They are running it as a real-time pricing engine — testing discount thresholds, term-length tradeoffs, and clause structures against win-rate and NRR data on a weekly cadence. The deal desk is now the highest-leverage operating function inside revenue, and the companies treating it that way are pulling 2 to 4 points of net revenue out of the deal book that competitors are leaving on the table.

What to Do in the Next 90 Days

The leadership response to contract-term compression is not a strategy memo. It is a tactical reset of the deal book, the renewal motion, and the financial model. Five moves are worth running in parallel inside the next quarter.

Rebuild the contracted ARR forecast off the new mix. Run the forward 24-month revenue floor against an assumption of 13% sub-one-year, 64% one-year, and 23% three-year — not the assumption the company was using in 2023. Most finance teams will find that the contracted floor is 8 to 15 percentage points lower than the board deck still shows. Better to know in June than discover at year-end.

Stress-test the prepay discount ladder. If the multi-year close rate has dropped while the discount stayed flat, the discount is no longer doing the work it was priced to do. Lower it. Move the saved margin into a renewal-pricing instrument — a locked rate, a usage credit, a future-product credit — that pays for buyer optionality more efficiently than the deeper prepay discount.

Move renewals onto a named-owner motion with a quota. If renewals are currently routed through CSMs as a default touchpoint, that motion is leaking money. Stand up a renewal pod inside revenue with explicit renewal-rate targets, a price book, and the same deal desk support a new-logo deal would receive.

Build mid-term outcome publication into the contract. Negotiate a quarterly written outcome report into every new annual contract, tied to specified usage and value metrics. This is the single move that most materially reduces buyer multi-year refusal — and it is harder for competitors to copy than a deeper discount.

Re-platform the deal desk as a pricing engine. Move it out of approvals and into experiment design. Test term length, prepay discounts, off-ramp clauses, and renewal terms against win rate and NRR with the same rigor a product team would apply to a paid-acquisition test. The deal desk is the new growth lab.

The Quiet Headline

The three-year SaaS deal is not literally dying. It is being repriced — by buyers who can now see a credible AI-driven scenario in which the vendor they sign today is the wrong vendor in eighteen months, by buying committees that have learned to negotiate optionality clauses into deals their counterparts in 2022 would have signed straight, and by CFOs who have decided that the option value of being able to walk is worth more than two-digit prepay discounts.

The companies that are going to win the next two years are the ones that stop treating contract-term compression as a procurement-side anomaly and start treating it as the new shape of the demand curve. The renewal motion is the growth motion. The deal desk is the pricing engine. And the board deck line called "contracted ARR" is going to need an honest update before someone else delivers it for you.

The vendors who get there first will look — to the market, to the board, and to their own buyers — like the ones who saw it coming. The vendors who get there last will look like the ones who tried to discount their way out of a structural change. By late 2026, the difference between those two groups will be visible in the multiples.