Agile Sounds Old? Think Again

Every management generation has its retired gods. Total Quality, Reengineering, Lean — each arrived dressed as a revolution, terrified the incumbents for a few years, and then slowly retreated into training decks and certification programs.

Agile joined that list faster than most: it began as a sharp, insurgent way of working and, in many organizations, ended up as a calendar full of ceremonies, a wall covered in artifacts nobody reads, and a lingering sense that whatever this was meant to be, it somehow became admin.

The word aged, but the need did not

While Agile was being politely escorted to the IT basement, something else walked through the front door and went straight to the work. AI did not wait for your operating model to be ready; it changed how quickly decisions are expected, how fast teams must learn, which skills suddenly matter, and which assumptions quietly expired while no one was looking.

That leaves a simple but uncomfortable contrast: technology is now moving like a sports car, while most operating models still handle like a bus. In that kind of world, Agile has not become obsolete; it has become invisible infrastructure — not the theatre on the surface, but the chassis underneath everything else.

When AI moves faster than your company

In most leadership conversations about AI, the story follows a familiar arc. Someone points to pilots in a few functions, a steering group that meets regularly, and working teams that are “exploring use cases.” There is visible activity, and the activity itself becomes the evidence that progress is being made.

The global data tells a more sobering story about what that activity is actually worth.

Recent large-scale studies of more than a thousand companies consistently find that only a very small minority — roughly five percent — are realising AI value at scale, while close to sixty percent report little or no material impact on their business so far, despite significant investment. Other research on AI transformation describes the same pattern from another angle, concluding that fewer than one in five companies have managed to scale their generative-AI efforts in any meaningful way; most remain stuck in experiments and pilots that never change how core work gets done.

In that sense, adoption is not the real problem; scaling is. Organizations have brought the tools into the building, but they have not redesigned the way they work.

This is why a line that looks dramatic on a slide is, in practice, simply descriptive: AI won. Now it is the operating model’s turn.

When executives talk about what still feels hard, they rarely mention “the model” explicitly. Instead, they talk about lead times that refuse to shrink, approvals that continue to climb long chains of signatures, teams that cannot adjust mid-quarter without triggering confusion, and governance that treats change as an exception to be controlled rather than the condition the organization lives in every day.

AI increases the tempo of decisions and learning, while the structure underneath continues to move at its old pace, and the collision between those two speeds is already visible in the work.

Strategies don’t change companies. Habits do.

Most leadership teams can now point to an AI strategy with some pride. There are immaculate decks, carefully defined priorities and use cases, risk matrices, investment envelopes, and position statements that have passed through all the right committees. The existence of the strategy creates a reassuring feeling: the future has at least been addressed.

The problem is that strategies sit on top of systems, and most of those systems were designed for a slower climate: annual cycles, tightly controlled information flows, and an environment that at least pretended to be stable. In that climate, a sophisticated AI strategy can create a very modern illusion — leaders feel prepared because the thinking is advanced, even while the organization remains largely unchanged.

What they actually need is not more intellectual readiness, but more practical capacity.

Capacity does not live in the document; it lives in what teams are allowed and expected to do on an ordinary Tuesday afternoon. It looks like autonomy that is more than polite delegation. It looks like cadences designed for iteration rather than ceremonial status reporting. It looks like structures that create enough coherence for hard decisions to be made close to the work, without every issue gasping its way up to the top of the hierarchy.

If you look closely at the transformations that do work, the balance of effort is very different from the way most organizations spend their time. Many practitioners now describe this as a 10–20–70 pattern: perhaps ten percent of the effort is about algorithms, twenty percent about data and technology, and the remaining seventy percent about people and processes — how work is structured, how decisions are made, and how change actually happens in practice. Most companies quietly invert that ratio and then wonder why the value never shows up.

From there, a sharper truth comes into view: you can have the most elegant AI strategy in your industry and still fail completely if your organization cannot learn fast enough to execute it. The organizations that will turn AI into real advantage are not the ones that think the deepest thoughts in offsites; they are the ones that learn the fastest from using it in the messy middle of day-to-day work.

The real bottleneck isn’t AI. It’s you.

One of the most useful features of AI is almost accidental: it behaves like a mirror.

Introduce AI into a slow, rigid structure and it immediately reveals the parts of the organization you have been quietly tolerating for years. Decision cycles that once seemed “a bit slow” suddenly look absurd next to machine-speed feedback. Silos that once felt manageable become visible bottlenecks. Misalignment that teams could previously paper over with goodwill and late nights now breaks entire workflows when the pace increases.

This is not just a matter of opinion; the economics are starting to show up. Some enterprises that have redesigned core workflows around AI — rather than sprinkling isolated use cases on top of old processes — are already reporting EBITDA improvements in the range of ten to twenty-five percent in those areas. The contrast with the majority, still experimenting at the edges while the core remains untouched, is stark. That gap is not primarily technical; it is about operating model and ways of working.

The diagnosis can feel harsh, but it is also oddly liberating, because it suggests that the problem is not a mysterious failure of technology. It means you can stop trying to fix structural issues with another platform, another pilot, another tool that leaves the underlying habits intact.

The real work is structural, cultural, and behavioral. It asks how teams are formed, how decisions are made, how learning is shared, and how power moves through the system. This is precisely the terrain Agile was designed to address before it was flattened into a checklist of ceremonies and roles.

Seen through that lens, it becomes difficult to escape a more honest statement: the disruptive force inside your company is not AI itself, but the mismatch between the speed of the environment and the slowness of your structure. The companies that pull ahead will be the ones whose ways of working can evolve at something like the pace of the tools they have already invited inside.

Culture: the operating system you can’t see but always feel

Beneath the visible layer of strategies, projects, and tools sits a quieter, more decisive force: culture.

Culture determines whether AI becomes fuel or noise. In a culture grounded in experimentation, transparency, and shared ownership, AI tends to find fertile ground. People try things, talk honestly about what happened, discard what does not work, and double down on what does. The organization does not just deploy a tool; it builds a learning loop.

In a culture organised around caution, rigid hierarchy, and the constant search for perfect information before any move, the pattern reverses. AI becomes another item to be approved, monitored, and contained. Curiosity is quietly treated as risk. Momentum is absorbed by recurring meetings with many attendees and very little authority in the room.

By now, the difference this makes is measurable. In one large-scale case, a global company that initially treated AI as a generic tool rollout saw usage stall at around twenty percent. When the same organization segmented more than one hundred thousand employees into distinct AI-related roles and designed tailored learning journeys for each group, adoption climbed to nearly ninety percent, and the gains showed up not only in productivity metrics but in people’s confidence using the tools. The technology was essentially the same; what changed was the way people were supported to work with it.

That is a more grounded way to describe what many like to compress into slogans: AI does not change who you are as an organization; it turns up the volume on whatever is already playing inside your culture. Where there is clarity, trust, and a habit of learning together, AI tends to amplify those strengths. Where there is confusion, fear, and ornamental governance, it amplifies that instead.

Culture, unglamorous as it can sound, shows up in very practical questions. Do teams have genuine clarity about what matters? Can they see one another’s work, or do they operate in parallel fog? Is there a real cadence for learning, or only a “lessons learned” slide at the end of a project that nobody wants to revisit? As the world speeds up, those questions stop being soft; they become operational.

The world will not slow down for your org chart

AI has introduced a level of unpredictability that is not going back in the box.

Capabilities appear in weeks rather than years. Workflows evolve in the middle of a quarter. Customer expectations are reset by events in completely different industries. The half-life of what you “know” continues to shrink, and your processes age faster than they were ever designed to.

This is not a storm you can wait out until conditions improve. It is the climate.

In that climate, the original promise of Agile — stripped of jargon, certification, and theatre — is almost disarmingly straightforward. Give teams enough clarity that they know what truly matters. Give them enough autonomy that they can act without waiting for a committee. Give them rhythms that force the work into the open so they can learn together while the environment continues to move.

At re.set, the Agile Management System (AMS) takes that philosophy and turns it into something you can actually run. It is an operating model in which strategy, structure, and culture are wired around continuous learning instead of annual performance theatre. It does not ask you to add more meetings; it asks you to create more coherence, so that the meetings you keep actually matter and the work between them is easier to see and to improve.

In a world built on relentless acceleration, business agility is less a buzzword and more a survival skill. It will not remove volatility or uncertainty, but it can teach your company how to move through both without tearing itself apart in the process.

Real agility is not a methodology; it is a way of giving your company a future.

Takeaways: What leaders can do now

Real agility does not live in slogans, posters, or town halls. It lives in what your teams are allowed and expected to do every week.

Here is where you can start — or restart — in the age of AI:

• Shorten the distance between insight and action.
  Replace long, brittle planning cycles with lighter monthly or biweekly reviews where priorities can genuinely move. AI shifts too fast for static roadmaps and annual commitments carved into stone.

• Make work visible across teams.
  One shared, honest view of the work beats twenty immaculate but isolated boards. When people can see how their work connects, duplication falls, misalignment shrinks, and AI use cases stop dying in the gaps between departments.

• Build multidisciplinary teams around outcomes, not departments.
  Squads organized around customer or business outcomes move faster than functional silos trading tickets and escalation emails. In this environment, structure is not just an org chart; it is a strategic choice.

• Normalize weekly learning loops.
  Short demos, shared insights, and rapid iteration are modest rituals that create an ambitious habit. The pace at which your teams are allowed to learn now sets the upper limit on how fast your organization can transform.

• Treat governance as a speed enabler, not a brake.
  Clear, simple guardrails let teams move quickly without fear. Vague, slow, or over-engineered governance guarantees escalation, delay, and a steady drift back to the status quo.

• Invest in culture with the same seriousness as technology.
  Psychological safety, curiosity, and accountability are not decorative concepts; they are the prerequisites for turning AI from an interesting experiment into compounding value.

• Align leadership around fewer, sharper priorities.
  Complexity is a quiet killer. When everything matters, nothing does. The discipline to choose — and to keep choosing — at the top creates focus and sanity everywhere else.

• Design your operating model for weekly change — because that is already your reality.
  Agility is not a dramatic reaction to disruption; it is the ongoing practice of building a company that assumes the environment will keep moving and chooses to move with it.

Closing Thought

Agile didn’t age; it was misunderstood. AI has made that misunderstanding impossible to maintain. The companies that thrive from here won’t be the ones with the most algorithms, but the ones capable of evolving at the speed their environment demands. Agility — the real, grounded, human kind — is what makes that evolution sustainable.