McKinsey Warns AI Vendors: Prove Value Before Raising Prices

A new McKinsey & Company report has cautioned software-as-a-service (SaaS) vendors against overhyping artificial intelligence without delivering measurable business impact. The consultancy’s latest analysis, “Upgrading Software Business Models to Thrive in the AI Era,” outlines how many vendors risk inflating costs for customers while struggling to prove tangible productivity gains. 

According to McKinsey, only 30% of AI software providers have published quantifiable return on investment (ROI) data from real deployments. Despite promises of automation and efficiency, many enterprise users report higher IT expenses — with AI-enabled tools driving up software costs by as much as 60–80%, without corresponding reductions in headcount. “All of these copilots are supposed to make work more efficient,” one Fortune 100 HR executive told McKinsey, “but my business leaders are saying they can’t reduce staff yet.” 

The report highlights three key barriers to effective AI monetization: 

• Lack of measurable savings: Many AI tools fail to demonstrate cost-cutting or productivity improvements. 

• Underinvestment in change management: McKinsey advises firms to spend $3 on training and adoption for every $1 spent on AI model development. 

• Unclear pricing models: Complex, usage-based billing makes it hard for customers to predict and manage costs. 

McKinsey suggests vendors adopt hybrid pricing models that blend per-user fees with consumption-based metrics, similar to Microsoft’s Copilot structure. However, it also warns that pricing strategies must evolve rapidly, as AI features that are “cutting-edge today can become table stakes tomorrow.” 

The consultancy notes that inference costs for large language models have dropped more than 80% annually over the past two years, urging vendors to align pricing with falling infrastructure costs and rising customer expectations. Still, with many enterprises reporting zero measurable AI ROI, McKinsey’s message is clear: AI hype must translate into real, defensible value — or risk a customer backlash. 

 

Source: 

https://www.theregister.com/2025/10/09/mckinsey_ai_monetization/