gshc2020.com

I have started companies during three major technology category shifts, so let me say what many leaders are thinking but not saying out loud: You are likely funding a transformation you cannot yet fully define, and so is everyone you benchmark against.

The numbers make the gap hard to ignore. Gartner projects enterprise AI spending will reach $2.59 trillion in 2026, a 47% jump in a single year. Yet IBM’s research finds only about 29% of executives can confidently measure the return on that spending.

The investment is real, but execution is lagging behind. Organizations are spending as though the destination is clear while still struggling to define what success actually looks like.

I've helped build three technology categories. This one feels different.

I've seen this pattern before with enterprise performance management and customer success. The budget arrives before the definition. Vendors rush in to supply the meaning. Everyone benchmarks against everyone else, even when few can clearly explain what they're benchmarking. That rhythm is familiar. What makes this moment different is what comes next.​

But I want to be honest about something, because it would be easy for someone with my background to pretend otherwise. The substance of this one is genuinely new, in a way the others were not. Performance management and customer success rearranged how people worked. This rearranges who does the work.

The last time technology didn’t just inform a task but actually performed the action, not only the analysis, was never. There is no prior cycle to pattern-match against for that. Not for me, not for the analysts, and not for the vendors selling you certainty.

So here is the distinction most leaders are collapsing, and it matters more than anything else in this article: The confusion is familiar. What we’re actually deciding is not. Money-arrives-before-the-meaning is an old story. But the thing the money is buying, the software that does the work rather than waiting to be used, is a genuinely new decision.

The instincts from a career of buying tools could mislead you here, because this is not a tool you use. It is work you hand over. Treating an unprecedented decision like a familiar purchase is the most expensive mistake available to you right now.

The real danger isn’t moving too slowly.

In an emerging category, leaders worry about being late. The more expensive risk is the quiet one: spending hard against a definition you didn’t write. When the category has no agreed meaning, the meaning gets supplied by whoever is selling. Every vendor will tell you that what they do is the transformation. Buy enough of them and you will have spent the line item without ever deciding what it was for.

The leaders who win a new category are never the ones who buy the most of it. They are the ones who define the problem for themselves before the market defines it for them.

That starts by asking a better question than the one most teams are asking. The question I hear most often is about tools. The question that actually matters is about work.

The common question asks, “What AI tools should we buy?”

The better question asks, “What work should our people stop carrying, and what should carry it instead?”

Answer the second, and the goal of the whole exercise comes into focus. It is not “having AI.” It is not even efficiency. It is capacity, growing what your existing team can carry, without growing headcount. That is the outcome a board funds and a CFO can measure. Everything else is sizzle.

Here are four questions to answer before you spend the line item.

If you are accountable for that budget, here is what I would settle before approving another dollar of it. None of these require a purchase. All of them protect the spend.

1. Define the goal before you shop.

Write down the outcome in plain language, capacity and the specific constraint you are trying to remove, before you take a single demo. A demo is designed to define the problem for you. Walk in with the problem already defined.

2. Separate what you’re experimenting with from what you’re committing to.

Some of this budget should fund cheap, reversible experiments (small bets you can kill next quarter). Some of it is a foundation you will build on for years. Treating those the same is how money gets wasted and stacks get stranded. Experiment freely at the edges. Commit deliberately to the foundation.

3. Insist on understanding what the AI is actually doing.

Before you trust a system to act, ask it to show its work, show you what it looked at and tell you what it is unsure of. The most dangerous output in enterprise AI is a confident, complete-looking answer with no visible boundaries. You, not the vendor, will answer to the board for the decisions made on top of it.

4. Write your own definition before the market writes it for you.

Decide what “AI GTM transformation” means for your business, the work it redraws, the capacity it creates and the outcome it serves, and use that definition to evaluate everything you’re shown. A definition you author is a strategy. A definition you inherit is a shopping list someone else wrote.

Ask this question to the C-suite.

There is no analyst report, no peer playbook and no vendor that genuinely has the answer of exactly what AI GTM transformation is. Everyone is improvising. In three years, the executives who look prescient won’t be the ones who bought the earliest. So it's up to us to define what it means for our organization.

Here is the honest question to take into your next budget review: Are you funding an AI transformation you have defined, or are you funding it because everyone else is?

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?