Is Your AI Spending Backwards?

Imagine two companies, each investing $10 million in AI. The first puts $7.5 million into technology and $2.5 million into the people and organizational changes needed to make it work. The second does almost the opposite: $4 million on technology and $6 million on the human side. Which one would you bet on?

If you’re like most finance leaders, your instinct probably favors the first. It feels more disciplined. The money goes into something tangible, something you can see and track. Yet according to BCG’s 2026 research, the companies leading in AI look a lot more like the second. Companies run by AI trailblazers devote about 60% of their AI budgets to training and upskilling. Everyone else devotes about 25% to the human side of AI transformation.

That gap suggests many of us are approaching AI with the wrong mental model. We treat technology as the real investment and the human side as a supporting cost. But the returns you’re counting on depend on changing how the business actually works. What if the biggest problem with your AI budget isn’t how much you’re spending, but where you’re spending it?

What You’re Really Buying

When a company invests in AI, it isn’t really buying technology. It’s buying an outcome: higher revenue, better margins, faster cycle times, smarter decisions. The technology makes new things possible, but nothing shows up in the financials until the business changes how it operates.

Think about a company that rolls out AI to make its sales team more productive. Account managers can now research customers, prepare proposals, and size up opportunities in a fraction of the time. The tools work, and people are personally more productive, but what happens to all that saved time?

If nothing else changes, probably not much, and the hours get absorbed into the day. To actually capture that value, someone has to rethink the way things work: processes, playbooks, expectations, and how managers lead. That’s real work, and it costs real money. Yet it rarely shows up in the business case that’s approved.

The question we should be asking is “What does it cost to change the business enough to capture the value AI makes possible?”

The Hard Part Comes After the Technology Works

Technology costs are comfortable territory for finance. We can get quotes, sign contracts, and track spending against budget. Changing how people work is messier. It takes time from managers and team members who are already busy running the business. It means redesigning processes, rethinking roles, adjusting incentives, and reinforcing new habits long after the launch euphoria is over.

Forecasting is a good example. Building an AI model that produces a more accurate forecast might take a few months. Getting sales, operations, and finance leaders to trust that forecast, change how they plan, and make different decisions because of it can take far longer. Anyone who has tried to change something in Sales knows exactly what I mean.

When we leave that effort out of the budget, we end up in a frustrating place. The technology is ready. The business isn’t.

When the Business Case Falls Apart

This is where it becomes a finance problem.

Picture an AI initiative approved at $2 million, with $1 million in expected annual benefits. The plan assumes the capabilities will be deployed within six months, with benefits ramping up soon after. But the effort to redesign processes and bring people along was underestimated. Adoption drags. The first year delivers only a fraction of what was promised.

The payback period stretches, the return shrinks, and leadership starts to wonder whether AI was worth it.

Nothing was wrong with the technology. It did exactly what it was designed to do. The investment fell short because the organization didn’t fund the work required to use it.

Now imagine if the same company had spent more upfront on process redesign, training, and change management. The initial number would have been bigger and perhaps harder to approve. But the benefits would likely have arrived sooner and lasted longer. Spending more doesn’t guarantee a better return. But spending too little on the things that create the return can undermine the entire investment.

I suspect that’s what the trailblazers have figured out.

The Question Behind Your AI Budget

Most AI business cases are thoughtful about the benefits and thorough about the technology costs. Where they tend to fall short is everything beyond the tech: the work of changing how the organization actually operates.

So before you approve your next AI initiative, it’s worth asking two questions. What has to change in how we work to capture the value we’re expecting? And have we truly funded the human side needed to deliver those changes?

Because the most expensive AI mistake may not be spending too much on technology. It may be spending millions on technology and too little on what it takes to make that technology pay off.

About Root Idea

Root Idea helps CFOs protect the business from AI risk. We work directly alongside finance teams to map AI influence, establish decision governance controls that hold up to board scrutiny, and deliver training and change management to make governance stick.

If your organization is scaling AI and governance hasn't kept pace, that's exactly the conversation we're built for. Learn more at rootidea.ai.