AI has entered the era of “demonstration.” The past 18 months of experimental frenzy with PoCs, pilots, and demos has given way to tougher questions from CEOs, boards, and investors. What is the impact? Why haven't we seen any revenue yet? Are we looking in the right place?
Most companies don't, according to AWS Innovation Strategists Saeed Hoda and Glenn Holland. And the problem isn't technology. That's the strategy behind it.
Hoda and Holland spoke at CFO Leadership's recent AI in Finance forum in a session dedicated to addressing this very issue. “Today is about building an AI strategy that’s worth building,” Hoda said. “What exactly does that mean? And what's stopping it?”
The answer, they argue, begins with a shift in thinking that CFOs have a special position as leaders.
AI strategies that start with AI are doomed to failure.
The first trap is also the most prevalent. So organizations start with technology. they ask, How can AI be applied to this? than that What problem is worth solving?
“The flaw is that there is too much focus on AI,” Hoda said. “It's not about using AI to do something. It's about making something better.”
But boardrooms often dictate the opposite. Leaders push “Gen AI-enabled” products not because the features improve customer outcomes, but simply because the label feels modern. “The number one reason AI projects fail is because the business case is not made,” says Holland. “They have no way of guessing what potential value they are going to get from these projects.”
This is where CFOs have to be assertive. Finance leaders have the clearest view of where customer friction creates financial resistance, where processes are inefficient, and where investments actually change outcomes. If AI isn’t connected to these realities, it won’t scale, no matter how smart the model is.
Healthy paranoia is now a requirement for leaders.
The second obstacle to a durable AI strategy is a lack of urgency. That is, a false sense of security inside what Hoda and Holland refer to as the corporate “castle.” Companies spend decades reinforcing what made them successful. Then another person appears with a helicopter.
“And all of a sudden, what you've built over the years starts to disappear because yesterday's customers have changed,” Holland said.
The helicopter moment is well known.
- Kodak invented the digital camera, but then suppressed it.
- Nokia dominated mobile phones until its OS collapsed overnight.
- BlackBerry believed its keypad was a moat until it wasn't.
As technology cycles accelerate, the helicopter moment won't be announced years in advance. They appear quarterly or every few months.
CFOs need to institutionalize that paranoia. Not fear, but caution. That means forcing leadership teams to answer difficult questions such as:
- Who will be the customers of tomorrow and what will they value?
- Who are the adjacent adversaries who can use AI as a weapon against us?
- Where are you vulnerable to speed, cost, and experience disruption?
Your AI strategy should be a portfolio, not a bet.
Even companies that foresee disruption coming often make the second mistake. It means being fully invested in today. While operational AI projects dominate budgets, initiatives that have the potential to redefine business are largely absent.
Holland and Hoda focused on an innovation framework created by IT service management company Gartner that directs three different types of initiatives. “The right portfolio has all three,” Hoda says.
- protect your project, “This basically helps us stay competitive.”
- Expand your project “This helps create new differentiations. Usually they are focused on increasing revenue and increasing market share.”
- turn over project, It's “about destruction. It's about changing the way we see good.”
But in many organizations, the tipping bucket that is most likely to future-proof the company is empty.
Holland pointed out that innovation portfolios are similar to venture capital portfolios, where outliers can generate huge profits. If CFOs allow long-term bets to shrink or disappear, companies effectively lose their futures.
And importantly, these categories cannot be measured using the same method. Hoda quoted a CTO who lamented, “My CFO wants to evaluate disruptive projects like defense projects.” But transformation projects won't deliver ROI this quarter or the next.
It is important that financial leadership becomes strategic leadership. CFOs must rebalance spending, rethink measurement, and give disruptive ideas enough oxygen to evolve. Because your competitors will too.
Holland joked that the CFO might need a second title: chief paranoia officer. Not because we should be afraid of failure, but because someone has to keep the organization on guard against failure.
“We truly believe that the fact that you already control some of the most important levers of your business makes you the best person to drive change, innovation, disruption and awareness within your company,” he said.
