You have more leverage than you think.
The data proves it.
What you pay for software depends on one thing: leverage.
In 2025, buyers gained it. Median deal sizes dropped as more companies showed up with real pricing context and pushed vendors harder, while seat compression increased.
But leverage is not evenly distributed. Some categories are highly competitive. Others are constrained by switching costs and vendor control. The outcome depends on where you’re buying.
This report breaks down what companies are actually paying, where buyers have leverage, and how that shows up in real deals.
The median cost of a new software purchase has fallen for two consecutive years. But in 2025, two things happened at once that make the story more interesting than a simple compression narrative: the typical software purchase became smaller, while large strategic deals became significantly larger. The floor dropped and the ceiling rose. And — for the first time in our dataset — companies bought more net-new software than renewed existing software based on transaction volume.
Annual Contract Value (ACV) is the annualized value of a software contract. Based on verified purchases in Vendr, January 2024 – March 2026. Average ACV can be heavily influenced by large deals; median reflects the typical purchase. Q1 = Jan–Mar · Q2 = Apr–Jun · Q3 = Jul–Sep · Q4 = Oct–Dec.
The cost of the “typical” new purchase dropped significantly in 2025 — median deal size fell roughly 43% year-over-year, from $11K to $6K. Renewal prices fell too, but far more modestly: only down about 8% from $27K in 2024 to $24K in 2025. Combined with new purchases overtaking renewals in volume for the first time, this points to buyers exploring more at lower price points, while generally retaining investments in what they already have.
The average annual purchase price tells a different story. While the typical deal got smaller, outsized platform and infrastructure commits pushed the overall average for a new purchase up sharply in the second half of 2025.
The net? Across the market, buyers are increasing investments at lower price points while also making larger strategic commitments.
Category deep dives, vendor profiles, timing data, negotiation levers, and where your leverage is greatest — all backed by 15,817 real transactions.
No spam. Just the data.
Three of the four top net-new purchases in 2025 were AI platforms and tools (Anthropic, Cursor, OpenAI). At the same time, the renewal list looks almost identical to 2024 (LinkedIn, Salesforce, Zoom). The tools companies are buying for the first time have changed significantly. The tools they’re renewing the most haven’t — at least not yet.
Ranked by verified purchase volume across all Vendr transactions in 2025. ✦ denotes AI-native platforms. New Entrants = first appeared on Vendr platform in 2025. Click any supplier to learn more.
Looking at broader categories, the macro story plays out very differently depending on what’s being purchased. Some categories saw double-digit renewal compression while others barely moved. The three charts below show the highest-volume categories in our dataset — chosen because they represent three distinct patterns in how pricing moved this past year.
More companies bought security tools for the first time last year than in prior years. New purchases in the Security category grew 75% by transaction volume from Q1 to Q4 2025 — the strongest growth of any category in our dataset. At the same time, renewal median fell 15% year-over-year — from $25K to $21K.
Renewals in the Sales category held flat. Here, we saw a renewal median of $37.4K in 2024 and $37.6K in 2025 — essentially unchanged, while most other categories saw some renewal compression. Another notable pattern is the Q1 spike and Q2 dip in average Annual Contract Value (ACV) visible in both 2024 and 2025. This highlights Salesforce’s dominant impact on the wider category — with a January 31 fiscal year-end, large Salesforce purchases pull the Q1 category average up sharply before normalizing the rest of the year.
The Collaboration category has the widest gap between what buyers pay to start and what they pay to renew. New purchases in the Collaboration category came in at a median of $5K in 2025. The median renewal was $21K — about 4 times higher. That’s the largest entry-to-renewal gap of any category we measured. That said, renewal median fell 8.6% year-over-year, so there is some downward movement at renewal.
This shows annual average contract value (ACV) by quarter for that category. January 2025 - March 2026.
The table below ranks categories by negotiation leverage — how much pricing flexibility buyers actually have when they show up prepared. The difference between “High” and “Low” isn’t subjective. It’s measurable in discount depth, deal structure flexibility, and how much competitive pressure moves the pricing.
| Category | Buyer leverage | Top levers |
|---|---|---|
| CRM | High | Competition, executive involvement, timing, growth, multi-year commits, and uplift removal. |
| APM / DevOps | High | Competition (New Relic, Dynatrace, Grafana), usage optimization, consumption rate card negotiation, multi-year commits — consumption-based pricing amplifies leverage for buyers who can forecast accurately. |
| Data Warehouse | High | Consumption-based pricing is the key lever — buyers negotiate rate cards and reserved capacity commits, not just seat counts. Usage forecasting accuracy directly drives savings. |
| Identity & Access | High | Competition — buyers who introduce alternatives and engage 90+ days before renewal typically see the strongest outcomes. |
| Cloud Security | High | Fast-moving category with aggressive competitors (Wiz, Orca, Prisma) — workload count and multi-year commits unlock significant discounts. |
| Sales Intelligence | High | Highly commoditized — ZoomInfo, Apollo, Cognism all compete on the same use case. Buyers who introduce two alternatives and negotiate on credit export limits (not just seat count) consistently land meaningful discounts below list. |
| Collaboration | Moderate | Competition, growth, early renewal, uplift removal — switching is painful but competitive pressure is real. |
| Project Management | Moderate | Build vs. buy pressure is real and Asana, Monday, and ClickUp all compete directly, creating strong timing and competitive leverage. |
| ERP | Moderate | Competition, growth, fiscal year-end timing, multi-year — sticky but negotiable. |
| Endpoint Security | Moderate | Competition between CrowdStrike, SentinelOne, and Microsoft Defender is real — endpoint count and multi-year are primary levers. |
| Design Tools | Moderate | Figma is more negotiable than it looks (competitive pressure from Adobe and Sketch). Adobe Creative Cloud is stickier and ecosystem lock-in makes switching painful. Annual commits with seat audits are your best lever. |
| Version Control | Moderate | GitHub and GitLab compete directly — switching is expensive but competitive pressure works. Multi-year commits and growth projections are your primary levers. |
| Compliance Tools | Low | Audit lock-in is real — regulatory requirements make switching painful. Multi-framework bundling at signing is your best lever. |
| Network Security | Low | High switching costs due to infrastructure integration. Multi-year commits and competitive quotes create modest pressure. |
| Social Media | Low | Commoditized, low-ticket category — budget constraints and uplift removal are the primary levers. |
Three factors determine where leverage lives:
1. Can you credibly build it? AI is shifting the “build vs. buy” calculus faster than low-code ever did. Companies are using Claude, ChatGPT, and Cursor to build internal dashboards, workflow automation, and knowledge management tools that would have required a SaaS purchase 18 months ago. When a buyer can credibly say “we’re evaluating whether AI can build this for us,” incumbents respond differently.
2. Is the market competitive? Identity & Access, Cloud Security, and Endpoint Security all have 3–5 credible vendors fighting for the same budget. Okta vs. JumpCloud. CrowdStrike vs. SentinelOne. Wiz vs. Orca. That creates negotiation leverage even when switching is painful. The overall pattern: buyers who introduce two alternatives and negotiate on per-unit rates (not just total price) consistently see better outcomes.
3. Is pricing consumption-based? Data Warehouse, APM, and Cloud Infrastructure buyers negotiate rate cards, reserved capacity, and commit discounts — not seat counts. Buyers who can accurately forecast usage and introduce competitive alternatives extract 20–40% better unit economics than those who accept the vendor’s first proposal.
Does build vs. buy pressure exist everywhere? No — at least not yet. With things like Compliance Tools and Legacy Network Security, audit requirements and switching costs shift power back to vendors. In that space, timing and multi-year commits provide your most meaningful leverage. CRM isn’t there yet either (Salesforce and HubSpot still own the enterprise market), but simpler categories that power project management, internal workflow automation, and internal dashboards are facing scrutiny.
The profiles below cover Salesforce, Datadog, Okta, OpenAI, and LinkedIn. Each profile shows when to engage, what moves pricing, and what contract terms to push back on.
Salesforce is the most negotiated vendor in enterprise procurement. The 9% standard uplift is the key variable. January is the single best month to negotiate — their fiscal year ends January 31.
| Term | 12 months |
| Payment | Net 30 annual (semi-annual/quarterly available, higher cost) |
| Pricing model | Seat-based, 9% standard annual uplift |
| Watch for | Premier/Signature Success, Own, Mulesoft, Slack, Tableau, user padding |
| Critical terms | No auto-renewal clause, 5% max uplift cap language, pricing schedule for growth commitments |
The data below comes from Ruth — Vendr’s AI pricing and negotiation agent. This rank is based on how many distinct buyers researched a specific supplier’s pricing during each quarter. High research activity suggests buyers are actively evaluating a vendor — whether for a new purchase, a renewal, or as a competitive alternative.
The quarter-over-quarter changes are the signal. When a vendor jumps in the rankings, buyer attention is shifting. When a vendor enters the top 10, it’s gaining traction as a benchmark. When a vendor drops out, buyers are researching it less — or they’ve already made a decision.
Anthropic’s jump from #10 to #2 is the steepest single-quarter rise in the dataset. This coincides with the ML & AI subcategory’s volatility in closed deals — median ACV ranged from $20K to $69K across 2025 as enterprise AI spend found its shape. Buyers are actively evaluating AI vendors, and that research activity is accelerating.
Buyers who landed 20–35% below list in 2025 knew three things: what comparable companies paid, when to engage, and what the vendor’s alternatives actually were. Buyers who paid median or worse didn’t.
It’s not negotiation skill. It’s not budget size. It’s data.
Most buyers still renew 60 days out instead of 90. They negotiate on total contract value instead of per-unit rates. They accept the 5% uplift because “that’s just how it works.” They don’t introduce competitive alternatives because they assume the vendor already knows they’re not switching.
The buyers who landed at the 25th percentile or better knew what comparable companies paid, when to engage, and what levers actually moved the pricing. The ones who paid median or worse didn’t. This report gives you the first part. Ruth gives you the rest.
She’s trained on $15B+ in verified spend data and 130,000+ negotiation outcomes. She knows what comparable companies paid, when they engaged, and what levers moved the pricing. And she can run the entire negotiation for you — or just give you the benchmarks so you can run it yourself.