A company can get faster at almost everything it does and still struggle to move forward. Reports arrive sooner, campaigns take less time to produce, developers can build more. Yet the customer still waits, the same decisions get stuck, and growth still depends on someone pulling five departments together.
There is a gap between improving the work and improving how the business works. AI is making that gap harder to ignore.
How much would we build again?
For much of the last few years, transformation meant working through the business we already had: moving services online, connecting systems, making data useful and reducing manual effort. There was substantial work in that, and there still is. But the possibilities have changed enough to reopen a fundamental question: how much of the current model would we build again?
An approval may exist because information used to be difficult to verify. A handoff may exist because two teams could not access the same knowledge. A customer may have to explain their situation three times because the organization divided its responsibilities that way. Technology can accelerate those steps. Understanding the business can reveal which ones no longer need to exist.
And there is an even bigger question: can we now imagine a different customer experience? Do we understand what matters most? Are there ways to unlock revenue that we could not have foreseen even yesterday?
This is where transformation, or reinvention, becomes even more interesting.
Understanding what holds the business together
The current model has a history. Someone introduced that approval after a costly mistake. Someone built that spreadsheet because the system could not answer an important question. Someone on the frontline quietly reconciles conflicting information every Friday so the customer never sees the problem. From a distance, these can look like inefficiencies. Up close, they may be holding the business together.
A credible transformation vision has to understand both: what the company could become, and what currently makes it function. Both now belong on our roadmaps, the possibilities we want to pursue and the foundations that make them achievable. Otherwise, we risk removing the workaround while leaving the problem it was solving.
That understanding takes work. How does the business make money? Why do customers choose it, hesitate or leave? Where does growth create pressure? What does the team know that has never made it into a process document? The answers change what is worth building.
Turning observations into a considered roadmap
A company trying to grow might appear to need more leads. Perhaps customers actually need clearer answers before buying. Perhaps operations cannot serve the demand profitably. Perhaps employees spend so much time finding information that little capacity remains for conversations that would win or retain business. The same technology could contribute to all three, but the business reasoning cannot be borrowed from one and applied to another.
It is also critical to distinguish a one-off from a pattern worthy of entering the redesign register. Savvy business users bring context. Human insight, supported by pattern recognition, connects observations and builds the list. A discipline anchored in the vision then assigns value, prioritizes and sequences.
Individual productivity gains do not automatically change this wider system. A six-month randomized study involving more than 7,000 knowledge workers found that AI changed activities people could adjust independently more readily than activities requiring coordination. Email work became faster; meeting time did not significantly change. That helps explain how people can become more productive while the organization operates much as before.
The competitive pressure reaches the whole model
If a competitor can reach the same customer through its presence in LLM-generated answers and recommendations, then serve them with fewer delays, better information and lower cost, isolated efficiency gains may not close the gap. The pressure reaches acquisition, service expectations and the economics of growth.
That is the reason to turn the model upside down: begin with what the business needs to deliver, then reconsider how people, decisions, information and technology come together. A customer conversation can reveal unmet needs and inform the next product decision. Knowledge previously held by one employee can become available across a workflow. Human judgement can be concentrated where it contributes most.
The opportunity becomes tangible when those changes improve the customer experience, create capacity or open a viable source of revenue. That gives transformation something concrete to deliver and measure.
Keeping the vision connected to delivery
In my latest projects, the speeding train has become even faster. Something difficult at the beginning can become feasible before the project is finished. Testing exposes a limitation; a customer reveals a need; a new capability changes the options.
Requirements and specifications become more important in these conditions. They make business intention precise enough for a team to act on together. “Help the customer” sounds clear until we decide what the system may promise, which information it should trust, what it can do and when a person needs to intervene. Those decisions carry through into testing, permissions, monitoring and recovery, the engineering discipline that makes the service dependable.
Reconnecting with the team keeps the work aligned. A developer discovers a constraint. A customer conversation changes the priority. Until those discoveries come together, people can be working diligently from different understandings of what matters. Re-evaluating the direction, updating the roadmap and carrying changes into delivery become part of the same discipline.
AI has expanded what we can imagine and build. Realizing that opportunity asks more of our business understanding, because we have more ways to act on it. Transformation earns its value when the business serves customers better, grows on sounder economics and can keep adapting.
The train keeps getting faster. The work of understanding where it needs to go deserves more attention than ever.
