David Craig’s line, “Systems don’t run companies, people do,” works because it corrects a common mistake in how leadership teams view change. Transformation is often described as a series of platform decisions, data initiatives, and AI deployments. The harder truth is that companies only change course when leadership clarifies what must be different, what structures block progress, and which people, standards, and decisions will define growth.
The Refinitiv story strengthens that argument. Carved out of Thomson Reuters, rebuilt at speed, and sold to the London Stock Exchange Group for £20bn, Refinitiv shows companies don’t reset just by modernising tools. Change happens when leadership faces what holds the company back and commits to what it must become. This makes Refinitiv’s story more interesting than typical transformation narratives; it’s about strategic separation, sharper choices, and leadership prepared to act when given the chance.
Competing through difference, not imitation
One of the strongest lines in the original interview is also one of the most useful for any leadership team operating in a mature market. “We wouldn’t try to out-Bloomberg Bloomberg. We chose to be different.” There is a broader business lesson in that decision because too many companies still confuse strategy with competitive mimicry. They study the category leader, copy its visible features, and assume that enough capital, effort, and technology will eventually narrow the gap.
That approach usually leaves a business reacting to another’s strengths. The incumbent has scale, customer embedment, and momentum. Trying to be a better version rarely creates value—especially in data and technology, where features are easily copied. The better move is to decide where to be different and organise around that position with conviction.
That question feels even more relevant now that AI is part of almost every technology strategy. When a powerful capability is widely accessible, advantage depends less on using the tool and more on the distinctiveness of the business. A vague market position does not become sharper simply because a company adopts AI more aggressively than its peers. Technology can enhance a differentiated model, but does little for a business that has not clarified why customers should choose it over established alternatives.
Some businesses do not need more optimisation. They need a reset.
The Refinitiv carve-out is revealing because it frames change as a structural issue before it becomes an operational one. The business had been competing within a framework that no longer suited what it needed to become, and the internal codename “Texit” captured that sense of release from a structure that had started to constrain both sides. The fact that Thomson Reuters’ share price later rose so sharply only reinforced the point that the existing arrangement had been limiting performance rather than supporting it.
That pattern is more common than executives admit. Companies try to optimise models that have run out of strategic room. They launch transformation programmes and new systems, but the real issue is higher-level: the ownership context, outdated models, or incentives that reward caution when markets demand decisiveness. More operational activity can look like progress, but leaves the constraint intact.
This is why preferring “reset” to “transformation” feels more honest. In the interview, Craig says the latter makes people “roll their eyes,” and it’s hard to disagree. Transformation has become a corporate catch-all that now means almost nothing. Reset suggests something more exacting. It implies that the old assumptions no longer suffice and that the business cannot just be polished into a better version of itself. A more accurate description of what is required.
People, not plans, are what make change real
“People, not plans, prove decisive” deserves more attention. It marks where strategy becomes real or stays theoretical. Large businesses have plenty of plans—strategy decks, workstreams, milestones, and governance. They often lack the alignment needed to turn plans into substantive change.
The hard part of change is not usually defining the ambition. It is creating an environment in which people understand what that ambition requires, believe leadership is serious about it and can see that standards are changing accordingly. This is where many programmes lose force. The organisation hears the language of change, but the incentives remain familiar, the tolerances remain loose, and leadership behaviour does not fully support the claims. In that environment, transformation becomes something employees are asked to talk about rather than something they are expected to deliver.
The Refinitiv example is useful partly because it does not pretend this process is comfortable. Performance management was intentionally demanding, and by Craig’s own admission, parts of the approach were “pretty brutal”. Not every organisation will choose to express accountability in the same way, and not every tactic should be copied mechanically, but the broader principle is difficult to dismiss. Serious resets require clearer standards, sharper judgment, and a willingness to confront underperformance before it becomes cultural drag.
Too many executive teams want the appearance of high performance without accepting the discomfort that creating it brings. They want agility without tension, accountability without difficult conversations, and stronger execution without changing what the organisation rewards or tolerates. His blunt instruction to “get the bad people before they get you” may seem severe, but the commercial logic is familiar to anyone who has seen weak behaviour spread unchecked in a business. Culture is shaped not only by who a company hires and promotes, but by what leadership allows to remain once it is clear that standards are slipping.
Strategy only works when the capital matches the ambition.
A major point in the interview warns that “a strategy is only as strong as the capital backing it.” It often gets overlooked but divides It often gets overlooked, but divides strategies that hold under pressure from those that fray as trade-offs become real. Teams still treat investor alignment as an add-on rather than an integral part it strategy. They create ambitious narratives about growth, repositioning, or reinvention, but often fail to test whether the ownership structure aligns with the investment horizon, risk profile, and operational disruption required to support those ambitions. When a gap arises, the organisation is caught between differing expectations: publicly, the business is building for the future, but in practice, it is steered by short-term pressures that constrain execution.
The result is strategic inconsistency: the company talks boldly of change but manages a cautious financial reality. This mismatch quickly becomes obvious to employees and customers, and is often exploited by competitors. Vision, resources, risk, and expected return must align for confident action. A strategy can sound compelling in the boardroom yet fail in the market if the capital to support it is not available.
Crises can accelerate what leadership already knows needs to change.
The advice to “never waste a crisis” The advice to “never waste a crisis” is often misused, but it contains an important point. But it removes institutional hesitation. Choices once viewed as awkward or difficult become unavoidable, and organisations that recognise the moment can move far faster than thought possible.
That was clearly part of the Refinitiv experience during the pandemic. The business had an early view of the impact through its China operations, which gave leadership a chance to respond while others were still underestimating how serious the disruption would become. The result, in Craig’s words, was that the company completed “10 years of digitalisation in just a few months”. That line matters because it reveals something many organisations prefer not to admit: the real constraint on speed is often not capacity, but the absence of a forcing event strong enough to override internal resistance.
Even then, the more important lesson is not about pace alone. It is about responsibility. In the interview, he reflects that it should not take a crisis to remind leaders that “looking after your team is the number one responsibility.” That point is easy to nod along with, but harder to live by when uncertainty rises, and operational pressure intensifies. A crisis can accelerate technology decisions and organisational change, but it also exposes whether leadership still understands that people are not a secondary consideration in a high-pressure moment. They are the business carrying it through.
Unique data matters more as AI becomes more common.
The part of the interview that feels most relevant to the current market is the argument that “data is becoming more valuable and more powerful – and unique data is even more precious.” That is the right lens for the AI cycle now underway because it shifts attention away from general excitement about models and back towards the commercial assets that actually create advantage.
There is a growing habit of talking about AI as though the technology itself will provide differentiation, but that logic weakens as tools become more widely available. In practice, the businesses that benefit most are likely to be those that already possess something scarce and difficult to replicate, whether that is proprietary data, domain expertise, customer embedment, or a strong workflow position. Public information and public models will only take a company so far. In markets where performance genuinely matters, the edge still comes from access to something others do not have and from the ability to organise, package, and distribute that advantage in a more useful way than competitors.
That is why the warning about AI and data foundations deserves to be taken seriously. “Jumping into AI without fixing those foundations is a huge mistake.” Many businesses are doing exactly that. They are racing to deploy AI into environments where the underlying data is fragmented, poorly governed and scattered across incompatible systems. From the outside, this can look like progress because the language is modern and the tooling is new. Inside the business, it often produces a more expensive version of the same underlying problem. A messy data estate does not become strategic simply because an AI layer has been added to it.
What the Refinitiv case also illustrates is that value does not reside solely in the data. It sits in the curation, packaging, delivery and trust that surround it. Much of the company’s information was public, but its commercial value came from turning that data into something useful in context and combining it with more exclusive datasets, creating a greater advantage. That is a much more grounded way of thinking about AI-era opportunity than the current tendency to treat every new model release as though it were a strategy in its own right.
What technology leaders should take from this
The more useful lesson here is not that every company needs a carve-out, a crisis or a hard-edged reset. It is that system’s investment that should never be allowed to stand in for strategic clarity. Leadership teams need to know where the business will be different, whether the current structure supports that ambition, whether the people model matches the standards being claimed and whether the capital behind the strategy is genuinely aligned with the path being proposed. Those questions remain more important than whichever toolset currently dominates the market conversation.
The temptation in every technology cycle is to believe that the next platform will compensate for ambiguity elsewhere in the business. It never does. Systems matter, and they matter even more in a world shaped by AI, automation, and data-driven decision-making, but they do not set direction, create accountability, build culture, or make hard choices when they become unavoidable. Companies are still run by people, and the businesses that understand that tend to build something stronger than those that assume better technology can substitute for leadership that is not yet clear enough about what it is trying to achieve.
