The SaaS pricing landscape has fragmented significantly since the early days of simple per-seat subscriptions. Enterprise SaaS vendors now use at least six distinct pricing architectures, and many use hybrid models that combine elements of two or more. Each model creates different commercial risks and requires a different negotiation approach. For the complete SaaS cost management framework, see our SaaS licensing complete guide.
This analysis is written from the perspective of former SaaS vendor pricing executives who now advise buyers exclusively. The specific mechanisms by which each pricing model extracts value from buyers — and the specific counters to those mechanisms — are drawn from direct experience on both sides of the negotiation table.
The Pricing Model Selection Isn't Neutral: SaaS vendors do not select pricing models arbitrarily. Each model is chosen to maximise the vendor's revenue from the specific usage patterns of their customer base. Per-user pricing maximises revenue when user counts grow. Consumption pricing maximises revenue when workloads scale. Understanding why a vendor chose a specific model tells you exactly where the revenue extraction pressure is — and where your negotiation leverage lies.
Model 1: Per-User (Named User) Pricing
Per-User / Named User Pricing
Per-user pricing is the most common enterprise SaaS model. Each provisioned user consumes one licence, regardless of how frequently they use the software. The model is commercially attractive to vendors because usage growth (adding users) directly and immediately increases revenue, while the cost to serve an additional user is negligible. The model is commercially dangerous for buyers because the incentive to provision users generously — "we'll add everyone just in case" — creates shelfware that auto-renews at full price indefinitely.
The average enterprise SaaS application on named-user pricing has a 28–35% inactive user rate. These are users who are provisioned and billed but have not logged in during the past 90 days. At $150/user/year for a 1,000-user deployment, a 30% inactive rate represents $45,000 annually in pure waste. Multiplied across a 50-application SaaS portfolio, the aggregate shelfware cost is material.
Negotiation Strategy: Conduct a usage audit 120 days before renewal — deprovision all users with zero activity in 90 days before submitting the renewal order. This forces the renewal negotiation to start from the reduced count rather than the inflated count. Vendors will often resist deprovisioning at renewal time; do it before the renewal conversation starts to establish the baseline.
Model 2: Concurrent User Pricing
Concurrent / Floating User Pricing
Concurrent user pricing bills for the maximum number of simultaneous active sessions rather than the total number of provisioned users. For applications where usage is intermittent — where users log in for specific tasks and then log off — concurrent pricing can be substantially more cost-efficient than named-user pricing. A 500-user organisation might only need 150 concurrent licences if peak simultaneous usage never exceeds 30% of the user base.
The risk is peak-period underestimation. When organisations size concurrent licences based on average usage rather than peak usage, they experience licence shortage events at periods of maximum demand — often during month-end, quarter-end, or critical business events. Some vendors implement "burst" charges for concurrent users above the contracted ceiling, which can create unexpected invoice spikes. Others simply deny access when the concurrent limit is reached, creating business disruption that generates internal pressure to purchase more licences at list price.
Negotiation Strategy: Analyse concurrent usage data for your peak periods — the 95th percentile of simultaneous sessions — rather than the average. Size the contract to cover the 95th percentile peak with a 10% buffer, and negotiate a contractual burst allowance (typically 10–15% above licensed concurrent count) for agreed notice periods rather than unlimited burst pricing at list.
Model 3: Consumption / Usage-Based Pricing
Consumption / Usage-Based Pricing
Consumption-based pricing bills based on actual usage volume — API calls, records processed, data ingested, compute consumed, or transactions executed. It is the most "fair" pricing model in the sense that you pay for what you use. It is also the most dangerous for budget management, because usage-based costs scale with adoption in ways that are difficult to forecast and easy to underestimate.
The specific risk in AI and data SaaS applications is that consumption costs grow non-linearly as use cases expand. An organisation that pilots an AI application with 100 users at $0.02 per query will find that scaling to 2,000 users for production use changes the cost dynamics entirely — both the per-query cost (if the vendor reprices at scale) and the total query volume (which grows faster than user counts as users integrate the tool into daily workflows).
Negotiation Strategy: Convert variable consumption exposure into committed spend commitments with volume discounts — essentially a pre-purchase of expected volume at a lower per-unit rate. Committed consumption discounts of 30–50% below pay-as-you-go rates are standard for annual commitments above $250K. Include a consumption cap or budget alert threshold in the contract to prevent surprise overage invoices, and negotiate overage pricing pre-agreed at committed-tier rates rather than list pay-as-you-go rates.
Model 4: Tiered Subscription Pricing
Tiered Subscription (Essentials / Professional / Enterprise)
Tiered pricing offers multiple subscription levels with progressively richer feature sets — commonly labelled Essentials/Standard/Professional/Enterprise or similar. The pricing differential between tiers is typically 2–3x, and the features that distinguish tiers are often capabilities that were present in the product from inception but are now gate-controlled to drive tier upsell.
The over-tiering problem is pervasive in enterprise SaaS. Our advisory data shows that in organisations with tiered SaaS subscriptions, an average of 40% of users are on tiers higher than their actual usage requires. This occurs because IT or procurement decisions are made at the organisation level ("everyone gets Professional") for administrative simplicity, even though a majority of users would function adequately on the basic tier. At a $50/user/month price differential between Standard and Professional tiers, over-tiering 300 users costs $180,000 annually.
Negotiation Strategy: Conduct a feature usage analysis before renewal — identify which tier-differentiating features each user cohort actually uses. Use this analysis to negotiate a split-tier contract (e.g., 200 Enterprise + 800 Standard) rather than a single-tier blanket contract. Most vendors will accept mixed-tier deployments when the total ARR is preserved. The savings from right-tiering often exceed the savings from straight-line price negotiation.
Model 5: Platform / Enterprise Flat-Rate Pricing
Platform / Enterprise Flat-Rate Pricing
Enterprise licence agreements (ELAs) and platform flat-rate licences cover a defined organisational scope (typically all employees or a defined business unit) at a fixed annual price, often with committed minimum terms of 3 years. The commercial logic is favourable for both parties at the point of signing: the vendor gets a large, predictable revenue commitment; the buyer gets a discount for the commitment and eliminates per-user provisioning administration. The problem emerges at true-up time.
ELA true-up provisions require buyers to pay for growth above the original contracted scope — typically annually or at contract expiry. When an organisation grows through acquisition, organic hiring, or technology adoption, the true-up can represent a significant unbudgeted cost. For organisations with active M&A programmes, ELA true-up exposure can reach tens of millions of dollars. Detailed true-up management guidance is available in our SaaS true-up guide.
Negotiation Strategy: Negotiate a fixed true-up rate pre-agreed in the ELA — specifying the exact price per incremental unit above the baseline, applied to the growth population. This eliminates the commercial uncertainty of future true-ups and prevents vendors from repricing growth at current list rates. For organisations anticipating M&A activity, negotiate M&A carve-out provisions — excluding acquired entities from the true-up calculation for a defined integration period.
Model 6: Outcome / Value-Based Pricing
Outcome / Value-Based Pricing
Outcome-based pricing ties fees to business results — revenue influenced, deals accelerated, risks avoided, or efficiency gains achieved. It is increasingly common in AI and analytics applications as vendors seek to justify premium pricing by tying it to demonstrable business impact. The concept is appealing — you pay based on value delivered — but the practical implementation is problematic because the measurement methodology is almost always defined by the vendor.
The core dispute in outcome-based pricing is attribution. When a vendor claims that their platform influenced $50M of revenue, the counter-question is: would that revenue have occurred without the platform? Attribution modelling for complex enterprise sales cycles is inherently imprecise, and vendors systematically apply models that maximise the attributed outcome. The financial exposure from outcome-based pricing is therefore significantly harder to cap or predict than any other SaaS pricing model.
Negotiation Strategy: Negotiate the measurement methodology, not just the price. Insist on contractual definition of the attribution model, the specific data inputs, and the calculation methodology before signing. Include a cap on total fees in any outcome-based arrangement — the cap should represent the maximum you would pay for the software on a straight per-user basis, so that the outcome model provides upside but not unlimited downside exposure. Require independent third-party verification of outcome calculations above a defined threshold.
Choosing Between Pricing Models
When vendors offer pricing model flexibility — which some do, particularly for enterprise-scale deployments — selecting the right model requires modelling your specific usage patterns against each model's commercial structure. Key questions: Is your user base growing, stable, or contracting? Is your usage concentrated among a core group of heavy users or distributed across a large, intermittent-use population? Are there significant usage peaks at specific periods? Is your use case well-defined or likely to expand?
For organisations with concentrated, high-volume usage: consumption pricing with committed volume discounts typically produces the best economics. For organisations with a stable, fully-deployed user base: named-user with aggressive shelfware management. For large enterprises with predictable headcount: platform/ELA with pre-negotiated true-up rates. For organisations with intermittent usage across a large population: concurrent licensing if available. For detailed negotiation strategy across the full SaaS lifecycle, see our SaaS vendor negotiation tactics guide.
Advisory firms including Redress Compliance and Atonement Licensing provide SaaS pricing model analysis as part of pre-renewal assessments — modelling your specific usage patterns against available pricing options to identify the optimal structure for each vendor relationship. See our SaaS License Optimization service for the full engagement scope.