{
  "id": 20890161,
  "title": "Why consumption pricing is breaking SaaS procurement",
  "url": "https://urgent.news/2026/11/20/why-consumption-pricing-is-breaking-saas-procurement",
  "topic": "tech",
  "section": "Tech",
  "published": "2026-11-20T10:34:21.000Z",
  "source": {
    "name": "TechRadar",
    "slug": "techradar",
    "url": "https://www.techradar.com/pro/why-consumption-pricing-is-breaking-saas-procurement"
  },
  "original_language": "en",
  "account": "The rise of consumption pricing in SaaS is causing procurement headaches for businesses. Uber's recent burning through its AI budget in just four months is a prime example of the challenges this new pricing model presents. Traditionally, software was sold on a per-seat subscription basis, but now 23% of SaaS vendors are shifting away from flat seat-based models in favor of usage, hybrid, and token-based pricing. This \"consumption-based\" pricing seems attractive on the surface, promising to align costs directly with business value. However, in practice, it benefits vendors more than customers. Contract values skyrocket when vendors introduce consumption pricing for AI features, leading to average contract value increases of over 50% for top tools. The illusion of cost control is deceiving, as switching to consumption pricing actually increases the average cost per user by up to 37%. The problem lies in the way consumption pricing monetizes every individual action, from CRM email drafts to automated data extraction. Each action incurs a small charge, incentivizing teams to experiment with AI tools and ultimately inflating the monthly SaaS bill. Traditional procurement levers are ineffective under consumption pricing. Average contractual discounts are 30% lower when paying by consumption compared to seat-based pricing. Vendors can easily argue that customers can simply consume less to reduce costs, trapping companies in a cycle of increasing expenses. Finance leaders struggle to model consumption costs, often focusing on deliberate human interactions while overlooking 35% of total consumption stemming from automated actions. As businesses move from human-prompted AI tools to fully autonomous AI agents, this idle meter will run continuously, consuming budget without direct oversight. Consumption pricing costs can surprise organizations by 15% within six months of signing a contract, far exceeding previous YoY budget variances of 4%. To control spending, finance and procurement teams must deconstruct billable units, define precise task costs, and establish the price of individual API calls, task summaries, or automated workflows. However, static procurement processes are ill-equipped to handle software contracts that act like volatile commodities. Implementing volume-tiered guardrails, capping overage rates, and negotiating dynamic pricing tiers are necessary to mitigate risks. Real-time visibility into cross-platform execution and spend is crucial to catch anomalies before they escalate into multi-million-dollar budget overruns. The transition to consumption and hybrid pricing in SaaS is becoming the new norm as AI capabilities expand across software categories. Vendors will continue pushing these models to align with their IT infrastructure costs and boost revenue. Businesses must recognize that managing dynamic financial models with static tools is ineffective. Unified visibility into daily spend patterns, usage context, and dynamic commitment tracking is essential to prevent software expenses from outpacing financial forecasts. Companies that align vendor pricing innovation with modern, data-driven approaches will be the ones that thrive.",
  "summary": "Consumption-based SaaS pricing causes unpredictable budgets, driving up software costs through uncontrolled usage.",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}