80% of companies investing in AI report no meaningful impact on EBIT. That is not an AI failure. That is a measurement failure.
Goldman Sachs published this finding in their 2025 AI Productivity Report and it sent shockwaves through boardrooms. But the data tells a more nuanced story than the headline suggests.
The Two Sides of AI ROI
The companies that measure AI ROI correctly are seeing extraordinary returns. McKinsey reports that top AI performers generate $3.70 in value for every $1 invested. Menlo Ventures found that enterprises with mature AI deployment see 40% reduction in operational costs within 18 months.
But here is the catch: only 20% of companies are measuring AI ROI systematically. The other 80% are either not measuring at all or measuring the wrong things.
What Most Companies Measure (Wrong)
The typical AI ROI framework looks at:
- Cost of AI tools and licenses
- Number of employees using AI
- Volume of AI-generated outputs
- Time saved on individual tasks
These metrics are meaningless without revenue context. Saving 30 minutes per day on email drafting does not matter if those emails do not generate pipeline.
What Top Performers Measure (Right)
Companies that actually see EBIT impact from AI measure:
Revenue Per AI Agent: How much incremental revenue does each deployed AI agent generate? The best companies track this weekly.
Cost Per Acquisition Delta: What is the before-and-after CPA when AI is deployed in the acquisition funnel? BCG reports top performers see 40-60% CPA reduction.
Pipeline Velocity: How fast do deals move through the funnel with AI-assisted selling versus without? HubSpot data shows AI-powered sales processes are 34% faster.
Customer Lifetime Value Impact: Does AI-enhanced service and personalization increase retention and expansion? Gartner found that AI-native customer success programs increase CLTV by 28%.
The N5R.ai AI ROI Framework
At N5R.ai, every engagement starts with establishing measurable baselines:
Week 1: Benchmark current CPA, pipeline velocity, conversion rates, and revenue per employee across all channels.
Week 4: First AI agent deployment with parallel tracking against baseline metrics.
Week 8: Full ROI assessment with statistical significance testing on all key metrics.
Week 12: Comprehensive ROI report with recommendations for scaling the highest-performing AI workflows.
Our clients do not wonder if AI is working. They know exactly how much revenue every AI agent generates.
The Bottom Line
If you cannot quantify the revenue impact of your AI investment within 90 days, you are doing it wrong.
Goldman Sachs is right that 80% of companies see no EBIT impact. But that is a strategy problem, not a technology problem. The 20% who get it right are building insurmountable competitive advantages.
Book a free 15-minute AI Agent Audit with N5R.ai. We will show you exactly how to measure and maximize your AI ROI.
