Stop Measuring AI ROI. You're Chasing the Wrong Number.

Antonio Mañueco · February 20, 2026 · 5 min read #ai #roi #strategy #consulting

Stop Measuring AI ROI. You’re Chasing the Wrong Number.

Hook

Every consulting firm is racing to build an AI ROI framework. Efficiency gains. Cost reduction. Time saved. Neat spreadsheets with percentages that make executives feel warm inside.

It’s a trap. And the firms chasing it are going to eat each other alive competing on a metric that doesn’t capture what AI actually does.

The Problem: Vanity Metrics Dressed as Strategy

Here’s how every competitor frames AI ROI:

  • “We saved 40% on document processing time”
  • “Our AI reduced support tickets by 30%”
  • “McKinsey deployed 25,000 AI agents” (and everyone laughed)

These are vanity metrics. They measure the shadow, not the thing casting it. They answer “did AI make the existing thing cheaper?” when the real question is “did AI make a different thing possible?”

That’s like measuring the ROI of electricity by counting how many candles you stopped buying.

The Uncomfortable Truth About ROI

ROI is a ratio: return over investment. Simple math. But it assumes you can isolate the return and attribute it to the investment.

With AI, you can’t. And here’s why:

AI Doesn’t Optimize — It Transforms

When Omar Armenteros said he’s working more hours with AI, not less, he broke the efficiency ROI model. His “return” isn’t fewer hours. It’s solving problems he couldn’t solve before. How do you put that in a spreadsheet?

When I built four platforms in one day, the ROI wasn’t “saved 3 weeks of developer time.” It was “created four new capabilities that didn’t exist yesterday.” The denominator is clear (my time + AI costs). The numerator is… everything that now exists that wouldn’t have.

The Compound Effect

AI’s real value compounds. The first insight leads to the second project, which reveals the third opportunity, which restructures the fourth workflow. Measuring the ROI of step one misses that steps two through ten only exist because of step one.

This is like measuring the ROI of learning to read by tracking how many books you bought. The return isn’t the books — it’s everything reading made possible across your entire life.

What Competitors Are Getting Wrong

The Efficiency Trap

Accenture, McKinsey, Deloitte — they’re all building AI ROI calculators. Input your headcount, your process times, your costs. Output: a projected savings number.

This works for AI as automation — the bottom tier. RPA with a better brand. But it completely misses:

  • AI as capability creation — things you couldn’t do before
  • AI as decision acceleration — better decisions faster, compounding over time
  • AI as organizational intelligence — the collective gets smarter, not just cheaper

The firms selling efficiency ROI are selling a commodity. The first one to undercut wins — and the margins collapse for everyone.

SEI’s Different Play: Weave ROI Into the Conversation

Here’s the positioning: we don’t hand you a calculator. We sit in the room with you and help you see what AI makes possible — not just what it makes cheaper.

The Workshop Approach

Instead of “here’s your ROI spreadsheet,” it’s:

  1. What decisions are you making today that you wish you could make faster? (Decision acceleration)
  2. What would you build if you had 10x your current capacity? (Capability creation)
  3. What data do you have that nobody’s using? (Latent value)
  4. Where are your people spending time on things AI should handle? (Yes, efficiency — but in service of freeing humans for higher work)
  5. What would your organization look like if every person had an AI partner? (Organizational intelligence)

The ROI emerges from the conversation. It’s not a number you calculate — it’s a vision you construct together. Then you work backwards to make it real.

The Three Horizons of AI Value

A framework that actually works:

Horizon 1: Efficiency (what everyone measures)

  • Cost reduction, time savings, automation
  • Easy to quantify, easy to commoditize
  • Table stakes — if you’re only here, you’re already behind

Horizon 2: Capability (what leaders are discovering)

  • New products, new services, new decision-making power
  • Harder to quantify, harder to compete with
  • “We can now do X” vs “We do X cheaper”

Horizon 3: Transformation (what almost nobody talks about)

  • Organizational identity shifts. Business model evolution. Market creation.
  • Nearly impossible to pre-quantify — you discover it by doing
  • “We became something different” vs “We did something different”

Most ROI frameworks only capture Horizon 1. The real value lives in 2 and 3. And you can’t spreadsheet your way there — you have to conversation your way there.

The Philosophy: You Can’t Measure What You Can’t Imagine

Wittgenstein: “The limits of my language are the limits of my world.” If your language for AI value is limited to efficiency metrics, your world of AI value is equally limited.

The ROI conversation needs to expand the client’s language for what AI can be. Not “how much do we save?” but “what do we become?”

This is also Maslow: you can’t self-actualize if you’re stuck measuring survival metrics. Efficiency is survival. Capability is growth. Transformation is self-actualization. Most firms are selling survival to organizations that need actualization.

Call to Action

Next time someone asks “what’s the ROI of AI?” — resist the urge to calculate. Instead, ask:

“ROI compared to what? The company you are today, or the company you could become tomorrow?”

That question is worth more than any spreadsheet. And it’s the conversation your competitors aren’t having.

← All articles