The 70% Rule: Why People, Not Technology, Determine AI ROI

There's a stat I quote in nearly every executive conversation I have right now: BCG's 2025 AI Radar found that top-performing organizations follow what they call the 10-20-70 principle. They allocate 10% of their AI resources to algorithms, 20% to data and technology, and 70% to people, processes, and cultural transformation.

Organizations that flip that ratio — spending the majority on tools while underinvesting in their workforce — consistently fail to generate measurable ROI.

If you've ever wondered why so many AI implementations look great on paper and then quietly underperform, this is most of the answer.

Why the technology side gets all the attention

It's easier to budget for technology. There's a vendor, a quote, a contract, a clear set of deliverables. You can put it in a procurement spreadsheet and a board update.

The 70% — change management, training, role redesign, communication strategy, resistance management — is harder to scope, harder to budget, and harder to justify until something goes wrong. So it gets cut. Or assigned to someone who already has a full-time job. Or replaced by a single training session and a Slack message.

What the 70% actually contains

The people side isn't one line item. It's a stack of practices, most of which have decades of research behind them. The ones I see consistently overlooked:

• Stakeholder analysis. Who is affected, how, and what do they need to hear?
• Communication strategy. Layered, repeated, role-specific messaging — not a single all-hands deck.
• Sponsor coalition. Active, visible, vocal sponsors at every layer of the organization, not just the C-suite.
• Role-based training. Different roles use AI differently. A finance analyst needs a different curriculum than a customer support lead.
• Resistance management. Naming the resistance, understanding its source, and addressing it directly — instead of pretending it doesn't exist.
• Reinforcement systems. Measurement, recognition, and follow-up that keeps adoption from drifting back to old habits.

"The technology was never the hard part. Getting your people ready is."

What it looks like when you get it right

The organizations getting real ROI from AI aren't the ones with the most sophisticated models or the biggest tech stacks. They're the ones whose people actually use the tools, trust the outputs, and have integrated AI into how they do their work.

That outcome doesn't happen by accident. It happens because someone treated the human side of AI rollout as a structured, fundable, measurable program — not a soft-skills afterthought. And it starts with the manager layer: when middle managers don't know what their new job is in an AI-powered org, the 70% stalls before it even reaches individual contributors. See The Manager's New Job in an AI-Powered Company for what that transition actually requires.

The bottom line

If your AI budget is 80% technology and 20% people, you are statistically likely to be in the group that underperforms. That's not a guess — it's what the data says.

The 70% is where the ROI lives. If you're not investing there, you're paying for a capability you won't use.

To see exactly what happens when organizations skip the people side, read Why Your AI Pilot Succeeded and the Rollout Failed. For the change framework that structures the 70%, see ADKAR for AI.