Enterprise technology conversations have shifted decisively over the past two years. The focus is no longer on whether organisations should adopt cloud, automation, or AI, but on how they turn those technologies into measurable outcomes. Across industries, leadership teams are discovering that access to technology is no longer the limiting factor. Execution is.
For Tom Soderstrom, AWS Executive in Residence, this change reflects what he sees consistently in discussions with senior leaders around the world. Having spent years on the enterprise side before joining AWS, Soderstrom now works with organisations navigating the transition from adoption to operational impact.
His perspective is shaped by experience rather than abstraction. “If you’re on the leading edge, you lead. If you’re on the bleeding edge, you bleed,” he says. “I was on the bleeding edge. I made some good things and some stupid things, but you learn from them.” Those lessons, he notes, increasingly have less to do with technology choices and more to do with organisational behaviour.
“When we talk to executives today, it’s rarely about technology,” Soderstrom says. “It’s about culture. It’s about the company culture, how they can move faster and implement things faster, and get their people trained faster.” The challenge is particularly acute in organisations that have grown large and successful. “If you were a small company that became successful, so now you grew larger and larger and older and older, you have a hard time innovating,” he adds.
At the leadership level, priorities have narrowed. Soderstrom says that in his conversations with CEOs and managing directors, one theme dominates. “The number one priority is speed — speed to market, speed to profitability, speed to compliance, speed to training and new skills.” Translating that urgency into execution, however, remains difficult in organisations built to minimise risk.
Why big bets slow everything down
One of the most persistent blind spots Soderstrom sees is scale. Under pressure to transform, many enterprises attempt change through large, multi-year initiatives. The result is often delay rather than momentum. “It’s very difficult to do something big, fast, and well,” he says.
Expectations around delivery have shifted. Where it was once acceptable to demonstrate results several years down the line, leadership teams now expect visible progress far sooner. The organisations making headway are not those lowering ambition, but those breaking it into smaller, measurable efforts.
This is where Soderstrom’s shorthand for execution discipline becomes relevant: think big, start small, scale fast. Thinking big establishes direction, but starting small anchors ambition in reality. “The starting small is the really important part,” he says. “You start with something that generates a business outcome that helps the company, that you can measure and then tell a story about.”
Those early results matter because they change behaviour. “That generates energy and excitement and the rest of the organisation comes along,” Soderstrom explains. Momentum, in his experience, is built through evidence rather than aspiration.
The mechanics of experimentation
Two operating principles underpin this approach. The first is working backwards from the desired outcome rather than leading with technology. “What is the business outcome I want?” Soderstrom asks. “And now I work backwards to try to figure out how do I get there?” Framing the problem this way ensures technology serves delivery rather than driving it.
The second principle is decision design. Too many organisations treat everyday choices as irreversible. “Most companies think that everything has to be decided by the top manager,” he says, which slows progress and concentrates authority unnecessarily. In reality, most decisions are reversible. “Most decisions are two-way door decisions. You can walk through the door. If this wasn’t worth it, you come right back.”
Cloud infrastructure makes this practical. “You don’t invest really anything. You just pay for what you use and come right back,” Soderstrom says. When decisions are small and risk is explicit, responsibility can be pushed closer to the people solving the problem, allowing organisations to move faster without losing control.
These dynamics are particularly visible in enterprise AI adoption. Over the past year, generative AI has moved from experimentation to board-level mandate, often accompanied by pressure to demonstrate rapid returns. Soderstrom sees this tension regularly. “The biggest challenge to AI is unrealistic expectations,” he says, followed closely by trust.
Trust operates on several levels. “There’s people-to-people trust. There’s people-to-software-agent trust. And then there’s agent-to-agent trust,” he explains. Without trust, adoption stalls regardless of technical capability.
Soderstrom’s advice mirrors the broader execution playbook. “Pick some real business problems, something small,” he says. “Pick something internal that’s not very risky.” Form small, cross-functional teams that include business, IT, and security, and set short timelines. “I used to give it two weeks, and they had to come up with a demo — something I can see.”
The goal is learning and adoption rather than polish. “Chief AI Officers, what they are really measured on in the end is adoption in the company of AI,” he says. Return on investment follows later. “Getting return on investment right away doesn’t happen,” but small investments and visible progress help reset expectations.
Concerns about talent inevitably surface alongside these discussions. Soderstrom challenges the assumption that organisations lack the people required to move forward. “You have the people you need,” he says. “They don’t have the skills they need.” Training, he argues, works best when it is tied directly to real work. “Don’t send them off to training and then they come back with nothing to work on. Give them on-the-job training.”
As organisations push further into AI and automation, the pressure to move faster is unlikely to ease. Boards will continue to ask for progress, returns, and evidence that investment is translating into advantage. What Soderstrom sees, however, is that the organisations making progress are not the ones chasing scale first, but those willing to experiment deliberately and adjust quickly.
“You start with something that generates a business outcome that helps the company, that you can measure and then tell a story about,” he says. “That generates energy and excitement and the rest of the organisation comes along.”
Discussion about this post