In June 2025, the European Commission reminded that while France possesses some of Europe’s strongest digital infrastructures, with widespread fibre and 5G deployment, the modernization of existing systems remains a major undertaking. Behind these solid foundations, many organizations still confront heritage architectures that slow down digital transformation.
The Illusion of Reinvention
In parallel with these constraints, modern architectures promise a new kind of simplicity. Since current best practices often start from a blank slate (migrate everything to the cloud, replace legacy languages with modern ones, adopt microservices, and set up continuous deployment pipelines), this can appear as a straightforward and reasonable solution, offering flexibility, scalability, and durability. On paper, it would enable teams to publish rapidly, fail fast, and adjust quickly. For many, it’s therefore the ideal: a system that invites change rather than resisting it.
Yet anyone who has lived through a complete rewrite knows that reality works differently. Projects expand, needs evolve, business logic recedes, and operational risk rises. Why? Because optimism collides with the scale of the legacy. What looks simple from afar becomes terribly complex up close.
It’s a bit like renovating a house and discovering plumbing running through walls intended for demolition. For organizations that have accumulated decades of code, data, and processes, a full rewrite quickly starts to feel like an illusion: ambitious, certainly, but rarely practical.
That doesn’t mean modernization is impossible. It simply must be approached differently. When executed well, modernization becomes tangible, reduces risks, and makes systems once deemed immutable more open to change.
The legacy then begins to earn its name: it adapts as much as it persists, and, as hard as it may be to believe, you only need to ask why these systems have endured for so long. The takeaway for IT leadership is that robustness, paired with evolvability, is a highly valuable asset.
Reframing the Narrative
Modernization does not pit the old against the new; it’s about weighing benefits and relative risks. A reinvented system can offer ultimate flexibility, but the dangers—such as outages, delays, costs, and erosion of organizational memory—are high.
Incremental modernization, from introducing APIs to adopting containers and CI/CD pipelines, can accelerate deliveries, improve integration, and ease developers’ onboarding. It may not deliver the sleek lines of a full rewrite, but it makes tangible progress for the organization. A small step in modernizing legacy often translates into a big leap in delivery.
The Reality of Legacy: Skills, Scale, and Fear
Legacy systems are hard to evolve for several reasons: skills are scarce, the scope is enormous, and they command attention. Nobody wants to be the one who breaks the machinery that has kept the business running for thirty years.
Conversely, some advocate “rip and replace,” a call to tear everything down and start from scratch, often without properly weighing the risks. It sounds bold, even visionary, but without prudence, that cry can become a eulogy: RIP—not only for the system, but also for the stability it guaranteed and sometimes for the CIO himself. Ambition without restraint, in the face of the real complexities of legacy, can quickly turn grand plans into career-ending moves.
Yet reducing legacy to a problem to be erased ignores its value: decades of business logic, embedded compliance, operational resilience, and optimized performance. These systems were not designed to change rapidly, but to run at scale with reliability. That does not make them obsolete. It makes them proven.
Mainframes, for example, continue to carry a large share of the world’s critical transactions. Far from being relics, they still offer a level of reliability, security, and throughput that many modern platforms strive to match. Rather than discarding legacy, the most durable path is to evolve it: make it observable, testable, and transformable. The goal is to leverage what works and modernize what slows the business.
Riding the Wave of Demands
Chief information officers must respond to a flood of requests. Internal teams demand more speed, product leaders want native digital capabilities, and auditors require compliance. At the same time, external voices grow louder: consultants urging a full rewrite, cloud vendors pushing for complete migration. “Move everything to us,” they repeat, as if the infrastructure alone would solve everything. Under pressure, the temptation to promise a sweeping transformation is strong.
Nevertheless, credibility is built through outcomes. A leader who demonstrates tangible progress—modernizing a key system, integrating cloud services, automating a manual process—changes the tone of the conversation. Frustration gives way to trust, and it then becomes possible to refocus on what becomes feasible, not merely what’s broken.
Added to this is regulatory pressure, which is equally structural. Measures like the Digital Operational Resilience Act (DORA) or NIS2 demand resilience, nudging CIOs to modernize not as a fashion, but for auditability, continuity, and control. This external constraint aligns internal teams, clarifies priorities, and opens the door to investments.
Placing It Within Strategy
Ultimately, surviving the legacy is not a matter of episodic tactics but of a strategic vision. The first step is to establish a precise current-state assessment: catalog what exists, quantify its business value, evaluate technical debt, and understand interdependencies. Then comes a road map, deciding what to keep, what to migrate, what to retire, and what to refactor. Finally, communication remains essential to align stakeholders, set realistic expectations, and highlight early wins.
The IT leader who aligns strategy with action builds trust. Over time, calls for a grand technological overhaul turn into opportunities for controlled progress.
In essence, the legacy is not the enemy, but it cannot be ignored. The IT leader’s role is to turn urgency into execution, ambition into architecture, and noise into momentum. Modernizing does not mean starting from zero; it means starting from what exists, with a clear destination in mind.
