💰TechCrunch AI•Jun 17, 2026Freshcollected in 44m

Enterprises struggle to justify AI ROI

Post LinkedIn

💰Read original on TechCrunch AI

#roi #enterprise-ai #cost-optimizationenterprise-ai

💡Learn why major enterprises are cutting AI budgets and how to avoid the 'tokenmaxxing' trap in your own projects.

⚡ 30-Second TL;DR

What Changed

Enterprises are re-evaluating AI budgets after overspending

Why It Matters

This shift indicates a cooling period for enterprise AI adoption, forcing vendors to prove tangible business value rather than just capability.

What To Do Next

Audit your current LLM usage and implement cost-tracking per project to identify and prune underperforming AI workflows.

Who should care:Enterprise & Security Teams

🧠 Deep Insight

Web-grounded analysis with 20 cited sources.

🔑 Enhanced Key Takeaways

•Many enterprises are currently stalled in generative AI pilot phases, with two-thirds unable to transition into full production and approximately 97% struggling to demonstrate tangible business value from these initiatives.
•The failure rate for enterprise AI projects is remarkably high, with various reports indicating that between 70% and 95% of these projects fail to deliver their intended value or measurable profit and loss (P&L) impact.
•A significant contributor to the poor ROI is the widespread use of activity-based metrics, such as token consumption, instead of outcome-based metrics directly linked to financial results like cost reduction, revenue growth, or improved employee experience.
•The total cost of ownership for enterprise AI is frequently underestimated, encompassing not only initial licenses but also substantial expenses for infrastructure, data engineering, specialized talent acquisition, continuous model maintenance, regulatory compliance, and complex integration with existing legacy systems.
•Challenges in generative AI implementation are often structural rather than purely technical, stemming from issues such as unclear project ownership, misaligned incentives, an inability to redesign workflows effectively, and inadequate data strategies.

🔮 Future ImplicationsAI analysis grounded in cited sources

Enterprise AI adoption will increasingly shift towards autonomous AI agents embedded in core business workflows, moving beyond basic chatbot functionalities.

The industry is observing a transition from simple generative AI tools to agentic AI systems capable of completing multi-step business processes, which offers a potentially clearer ROI model by comparing against fully loaded employee costs.

AI vendors will increasingly adopt outcome-aligned or usage-based pricing models, moving away from traditional per-seat licensing structures.

The current difficulties in demonstrating ROI from fixed licenses and the unpredictability of token-based costs are compelling vendors to offer pricing models that scale with actual business results or usage, aiming for more predictable costs tied directly to value.

Enterprises will prioritize the development of robust data strategies and comprehensive governance frameworks to effectively unlock AI ROI and mitigate associated risks.

Data quality, privacy, security, and seamless integration with existing systems are identified as critical barriers to achieving AI ROI, necessitating a strong focus on foundational data strategies and governance for successful AI implementation.

⏳ Timeline

2014

Google acquires DeepMind, signaling significant early investment in AI by major tech companies.

2017

McKinsey's Global Survey on AI reports a doubling of AI technology adoption rates, with 50-60% of companies increasing their AI investments.

2023

Generative AI gains significant traction, leading to an explosive growth in private investment, peaking at $25 billion, up from $3 billion in 2022.

2024

Corporate AI investment reaches $252.3 billion, with 78% of organizations reporting AI use in at least one business function, largely driven by generative AI deployments.

2025

MIT's NANDA report reveals that 95% of enterprise AI pilot programs delivered zero measurable P&L impact, highlighting a substantial gap between investment and realized value.

2026

The 'tokenmaxxing' trend becomes a prominent management concern, leading companies like Amazon, Microsoft, Meta, and Uber to move away from usage-based AI leaderboards due to inflated costs and a lack of alignment with business outcomes.