top of page
What your AI Budget Leaves Out

What Your AI Budget Leaves Out.

Why the cost of change belongs in the AI business case

Every AI investment rests on one belief, whether leaders say it out loud or not: people will change how they work. The tool creates value only when they do. That makes people the real subject of an AI business case, and the cost of their change belongs in the budget.

​

Picture a COO at a mid-sized manufacturer. She's reviewing next year's AI budget, and the lines are clear: software licenses, integration with the ERP system, data cleanup and a two-day training session from the vendor. The business case shows faster reordering and less excess inventory.

 

Six months after launch, a few buyers use the tool's reorder suggestions every day. Most still order the way they always have, from experience and a spreadsheet they trust. The vendor confirms the tool works as designed. What was missing was never in the budget.

​

What the budget usually covers

​

AI budgets are good at capturing what comes with an invoice. Licenses, integration, data work and a first round of training each have a price, a vendor and a contract. They're real costs, and they're necessary.

​

They're also the smaller part of the work. Boston Consulting Group uses a rule of thumb drawn from its client work: about 10% of the value from AI comes from algorithms, 20% from technology and data, and 70% from people and processes. It's a consulting heuristic, so treat the exact split loosely. The direction is what matters. Most of the effort sits where most budgets spend the least.

​

What it leaves out

​

Four costs rarely show up as line items.

​

Time to learn. People need time to try a new tool, make mistakes and build confidence, on top of their regular work. In BCG's 2025 AI at Work survey of more than 10,600 workers, only 36% said they felt adequately trained in AI.

​

Managers' time. Managers answer questions, coach hesitant team members and decide what to do when the tool and the old way disagree. That work happens in the margins of an already full day.

​

Redesigned work. When AI takes over part of a task, the task changes. Roles, handoffs and procedures need to be rewritten. In regulated settings, controlled documents and validations change too.

​

Decision rights. When AI suggests a reorder quantity or a reply to a customer, who decides whether to accept it? Who answers when it's wrong? Settling those questions takes leadership time.

​

When these costs go unfunded, people still pay them, informally. They build workarounds and keep the old spreadsheet "just in case." In the same BCG survey, regular use of generative AI among frontline employees had stalled at 51%, well below leaders and managers. Over time, those workarounds add to what I call Transformation Readiness Debt™, the unfinished change an organization carries from one initiative to the next.

 

What a complete budget gives you

​

When leaders fund the change itself, the payoff shows up in daily work.

  • People use the tools you paid for. BCG found regular use was sharply higher among employees who reported at least five hours of training. Five hours per person is a small line item for that difference.

  • Workarounds fade. The spreadsheet "just in case" stops being necessary once people trust the new way.

  • Decisions move faster. When people know where AI recommends and where a person decides, recommendations turn into action.

  • You can see whether it's working. Measuring adoption after go-live shows where the investment is paying off and where it needs attention.

 

What leaders can do

​

Put change costs in the business case. List them next to the technology costs: learning time, managers' time, work redesign and the leadership time needed to settle decisions. Even rough estimates make these costs visible and turn them into deliberate choices.

 

Fund managers explicitly. If managers are expected to coach their teams through AI, give them the hours and the skills to do it.

 

Settle decision rights before launch. Decide where AI recommends and where a person decides. Name who answers for the outcome, and write it down.

​

Measure adoption past go-live. Track whether people use the tool and whether they trust what it produces. Keep part of the budget open to act on what you learn.

​

In my work, these costs sit inside what I call Make Room. Structural room means clear decision rights and accountability. Psychological room means time and safety to learn. Both cost something, and both protect the investment.

 

The leader your people will follow

​

A budget shows what a leader values. One that names learning time, managers' time and clear decisions tells people their success with AI matters.

 

Leaders who fund that work earn teams that trust them with the next change. And when the board asks how the AI investment is going, they have the answer every executive hopes to give: people use it, and it's paying off.

 

A question for your next budget meeting

​

Before the next AI investment is approved, ask: what will it take for our people to actually work this way, and have we funded it?

This is the work behind my approach, Make Sense. Make Room. Make It Ours.™ You can see how it works on the My Approach page at adroitta.com.

 

Research references

1. Boston Consulting Group. The Leader's Guide to Transforming with AI. https://bcg.com/featured-insights/the-leaders-guide-to-transforming-with-ai

2. Boston Consulting Group. (2025). AI at Work 2025: Momentum Builds, but Gaps Remain. https://www.bcg.com/publications/2025/ai-at-work-momentum-builds-but-gaps-remain

Maria Isabel Meinholz
AI Leadership & Cultural Adoption Advisor | Founder, Adroitta

​

Bring your people with you through AI​

bottom of page