Summary

  • Christel Heydemann's consequential contribution was to make copper decommissioning in France part of Orange's declared network-modernisation programme, with a 2030 horizon, and later to describe the shift to full fibre as an industrial project. Those are executive choices and commitments, not evidence that she personally carried out local migrations (Orange's 2023 strategic-plan announcement; Orange's later retrospective).
  • The true unit of execution is the commune assembled into an annual lot. Selection, a stakeholder-sharing phase, formal notice, commercial closure and technical closure convert a national destination into local decisions that can be challenged, revised or postponed (ARCEP on Orange's closure organisation).
  • Fibre coverage alone is limited public evidence. ARCEP's framework couples complete fibre infrastructure with wholesale and retail offer availability, non-discrimination, notice periods, information sharing and a duty to postpone when closure criteria are not met (ARCEP on Orange's obligations through the closure phases).
  • Migration is a distributed operating responsibility: Orange steers closure of the network it owns; fibre-infrastructure operators make replacement infrastructure usable; commercial operators manage customer relationships and migrations; local authorities identify public assets and harder-to-reach users; and professional users inventory and replace copper-dependent services (ARCEP on closure organisation; ARCEP's local-authority guide; ARCEP's professional-user guide).
  • The operating compact is credible precisely because it does not equate an announced date with readiness. Data files expose gaps, local consultation surfaces constraints, and postponement preserves continuity where the regulated preconditions are missing (ARCEP on closure obligations).

Switching off is a different managerial act

Telecommunications strategy often describes deployment as an accumulation: more premises made connectable, more fibre in the ground, a larger footprint. A shutdown reverses the grammar. Its decisive question is not what has been added but what still relies on the system being withdrawn. A copper line may carry ordinary fixed telephony or broadband, but it may also sit beneath a lift call system, a payment terminal, an alarm, a public building or a link used to alert emergency services.

At technical closure, a dependency that has not moved does not become slower or less fashionable; the service stops. ARCEP's professional guidance states the consequence plainly: an enterprise still on copper at the technical deadline loses the connection and the services resting on it (ARCEP's professional-user guide).

That asymmetry explains why fibre investment and copper retirement cannot be treated as the same programme viewed from opposite ends. Deployment can report infrastructure availability while leaving the final customer connection, contract, terminal and power arrangement unresolved. Closure must trace the whole chain. ARCEP defines a fibre-ready premise for the infrastructure criterion by reference to the optical connection point being installed; that does not mean the final connection to the subscriber has already been completed.

The regulator separately requires the relevant wholesale and retail offers to be available before commercial closure. In other words, civil infrastructure, a marketable service and an actual user migration are distinct states (ARCEP on closure obligations).

The distinction creates a demanding form of executive accountability. In February 2023, Orange's Lead the Future announcement put decommissioning France's copper network by 2030 within a broader commitment to modernise fixed networks. Heydemann, identified there as chief executive, attached the plan to three guiding ideas—performance, excellence and trust—and invoked the commitment of Orange employees to its execution (Orange's 2023 strategic-plan announcement).

The pertinent choice was to treat withdrawal as part of modernisation, rather than allowing fibre construction and copper operation to continue as indefinitely parallel worlds.

But an executive declaration cannot itself satisfy a fire-safety obligation, activate a wholesale offer or find an overlooked analogue line in a municipal building. The useful way to assess Heydemann's role is therefore narrower and more exacting than personalising every result. She can be held to the public design and priority: a time-bounded exit, described as a major industrial undertaking, whose execution had to be organised across the company.

Orange's own governance description places the executive committee she leads in charge of implementing strategic orientations, overseeing major transformations, risk and resource allocation (Orange's governance page). Operational credit and fault, however, must follow the actors ARCEP names at each stage.

The executive choice: make retirement an industrial programme

Orange's 2023 announcement matters because it joined a technological destination to an end date. The company said it would continue deploying, operating and marketing fibre while decommissioning copper in France by 2030 (Orange's 2023 strategic-plan announcement). Those verbs describe overlapping obligations. During the transition, the replacement network must expand and mature, customers must be offered a route across, and the old network must remain serviceable wherever users still depend on it.

Retirement is not a single future event; it is a period in which two systems and their constituencies have to be managed together.

Heydemann's 2023 account, published by Orange after the first year of Lead the Future, said the group was mobilised around execution and singled out the broad extension of very-high-speed access and fibre as a significant result. She also called the progress collective (Orange's 2023 integrated-report interview). That is useful evidence of her public managerial framing: mobilisation, implementation and shared effort. It is not independent proof that any municipality or customer migrated successfully.

Corporate testimony can show what leadership claimed to prioritise; the regulator's record is needed to see how closure was structured and where readiness controlled the schedule.

A later Orange statement provides a retrospective bookend, but it needs the same discipline. In February 2026 the company said the 2023–2025 plan had met its objectives, while Heydemann described France as moving to full fibre through implementation of a copper-decommissioning industrial project (Orange's 2026 retrospective). The relevant phrase is “industrial project”: it confirms that the company regarded retirement as an operating system, not merely a regulatory date.

The claim that the preceding plan succeeded remains Orange's own assessment, not an independent verdict on continuity in every location.

The strongest leadership inference is consequently about institutional design. By putting retirement inside the strategic plan, management made unresolved migration a programme issue rather than leaving it as scattered customer service. By describing the work as industrial, it accepted repetition at scale: each lot would need comparable data, notices, market checks, migration work and escalation. And by making the destination public, it created a schedule against which ARCEP, rival operators, local authorities and users could organise.

None of that makes the chief executive the local operator. It makes her accountable for whether the organisation treats local exceptions as information to be managed rather than noise to be wished away.

This is where the operating compact begins. It is not a bargain in the sense of a private agreement negotiated once. It is a recurring allocation of duties under public rules. Orange chooses and steers the closure perimeter because it owns the copper network. ARCEP sets conditions and can examine compliance. Fibre-infrastructure operators must make the alternative network coherent with the retirement sequence. Commercial operators must move their own customer relationships. Municipal actors bring knowledge of territory and public services.

Users and their specialist suppliers must identify terminal-level dependencies. The executive contribution is to make that many-sided system executable inside Orange; the system's legitimacy comes from responsibilities that do not collapse into Orange's timetable alone (ARCEP on closure organisation; ARCEP on closure obligations).

The lot turns a national ambition into a governable risk

The pivotal design choice was geographical and procedural. Orange organised technical closure in annual lots, using the commune as the principal operating unit. ARCEP explains the rationale attributed to Orange: the commune makes communication to households and businesses more legible and helps involve elected officials. Orange establishes candidate lots using criteria that include fibre coverage and balance among operators and territories, with the criteria able to evolve in light of experience from early lots (ARCEP on closure organisation).

The commune is not technically pure. Networks cross administrative boundaries, operators have different footprints, and premises within the same place can face different connection conditions. Its advantage is institutional. A mayor can recognise public buildings and local groups; a municipality can inventory its contracts; commercial operators can map customers to a deadline; and an infrastructure operator can expose address-level holes within a perimeter local actors understand. The lot then aggregates those units into a repeatable wave.

National scale comes from giving heterogeneous places the same sequence of gates, not pretending France is homogeneous.

Selection is therefore an exercise in risk composition, not a league table of fibre percentages. ARCEP's account of the applicable criteria says Orange considers the fibre level, the representation of fibre-infrastructure operators in a lot relative to their wider footprint, the possibility of recovering equipment needed to maintain legacy telephone services, and the volume of copper accesses still active in a commune. ARCEP notes that this list is not exhaustive (ARCEP on closure obligations).

The mix matters because a lot must be large enough to industrialise but not assembled so narrowly that one operator, one territorial type or one easy migration case determines the result.

Before finalisation comes a sharing phase. Orange circulates the proposed communes so operators, local authorities and ARCEP can identify obstacles. ARCEP records that comments on early lots led to communes being removed for reasons including a mayor's or delegating authority's request and difficulties with local fibre deployment. During that process the regulator monitors non-discrimination among operators (ARCEP on closure obligations). This is an important correction to the idea that consultation is ceremonial.

It operates before the lot hardens, at the point when local evidence can still change the perimeter.

Once stabilised, the lot is notified to operators and the affected authorities with commercial- and technical-closure dates. That act starts the formal notice periods (ARCEP on closure obligations). The sequence creates a useful separation of powers. Orange proposes and officialises because it controls the asset being retired. Other parties test the proposal against deployment and migration reality. ARCEP supplies the conditions within which the dates become operable.

Local authorities are not asked to approve national network strategy, yet they are given a defined moment to surface territorial impediments.

The experience of 2024 and 2025 shows the mechanism doing more than announcing. ARCEP records that the first full lot closed commercially in January 2024 and technically at the end of January 2025, while three communes had their technical closure moved to June 2025. It also records that, for the second lot, Orange announced commercial closure in 765 communes in January 2025 while postponing it in 64 where compliance with the closure criteria could not be assured on time (ARCEP on closure organisation).

These facts do not establish flawless migration. They establish something more useful for evaluating the compact: the perimeter and date could yield to readiness.

Two closures, and the interval that makes migration possible

The words “copper shutdown” hide two legally and operationally different moments. Commercial closure stops the sale of new copper access in a defined area, regardless of which operator would have supplied it. Existing copper services continue. Technical closure later ends the services running over existing copper subscriptions, again across operators (ARCEP on closure organisation). The first changes the direction of market flow; the second removes the old production platform.

That distinction is the programme's central piece of sequencing. Once new sales stop, operators no longer add fresh demand to a network with a retirement date. Yet customers already on copper retain a migration window. Commercial closure thus works like a one-way gate: it prevents replenishment of the installed base while leaving time to unwind it. Technical closure is not a stronger version of the same commercial decision. It is the moment at which remaining dependence has an operational consequence.

ARCEP's default rule requires 36 months' notice before a zonal commercial or technical closure and an irreducible interval of 12 months between the commercial and technical milestones. Where fibre deployment is already particularly advanced when a lot is made official, the notice can be shortened under the regulator's framework.

The purpose of notice is expressly two-sided: alternative operators need time to deploy active equipment where they intend to be present and to organise migrations, while end users need time to select an offer and complete any necessary works (ARCEP on closure obligations).

Notice is therefore productive capacity. It gives commercial operators a horizon for contacting customers and gives fibre-infrastructure operators a horizon for resolving incomplete addresses. It lets a business find lines outside its obvious internet contract and lets a municipality place migration work within procurement cycles.

It also allows the copper network to remain a continuity instrument until the final gate: ARCEP stresses that copper service quality remains important where fibre is absent and households or businesses still depend on the old network (ARCEP's local-authority guide). Running the old network during transition is not strategic indecision; it is part of an orderly exit.

The sequencing also clarifies what a date can and cannot prove. A commercial-closure date shows that no new copper subscriptions should enter the zonal stock. It does not show that every existing service has migrated. A technical-closure date expresses the point by which migration must be complete, subject to the regulatory conditions and postponement process. It does not by itself prove that every terminal, organisation or public service was discovered early. Good governance depends on preserving these distinctions in reporting.

Conflating the milestones would create false assurance at commercial closure and dangerous lateness at technical closure.

Wholesale access is part of continuity, not an operator-side detail

A complete fibre footprint can still fail as a replacement market. Premises may be technically connectable while a customer's chosen retail provider lacks an active presence, an appropriate wholesale input or an offer suited to the use. ARCEP's framework therefore makes the availability of wholesale services between operators and retail services between a commercial operator and the user a closure criterion.

Orange must demonstrate before commercial closure that offers are available for households and professional premises, responding to end-user needs on technical and pricing conditions comparable to those received on copper (ARCEP on closure obligations).

This requirement changes the economics of shutdown. If ownership of copper allowed Orange simply to withdraw the old input before rivals could serve customers on fibre, retirement could distort competition while appearing technically complete.

The non-discrimination obligation prevents Orange from favouring one fibre-infrastructure operator over another during lot formation, while the notice period gives alternative commercial operators time to install active equipment at the relevant shared points and prepare customer migrations (ARCEP on closure obligations). Readiness thus includes contestability: the replacement network must support an operator ecosystem, not only an engineering route.

The rule also disciplines Orange's dual position. It is the owner initiating and operationally steering copper closure, but it is also one of the operators serving end users. ARCEP assigns commercial operators collectively a key role in end-customer relations and organising migrations, while fibre-infrastructure operators—including local authorities where public-initiative networks are involved—must align fibre deployment with the closure plan (ARCEP on closure organisation).

This allocation prevents “Orange” from becoming an imprecise answer to every question. Its network-owner duties, its retail duties and the obligations of other actors are related but not interchangeable.

Wholesale readiness is also a continuity safeguard for specialised demand. A business replacing several copper lines may not need a one-for-one set of fibre lines; it must first identify the services and resilience characteristics required, then ask commercial operators what configurations they offer. ARCEP notes that some uses may require multiple fibre accesses and directs users to consult operators present at the address (ARCEP's professional-user guide). The economic question is not simply whether “fibre” exists.

It is whether a service with the required characteristics can be ordered, installed and supported.

This helps explain why retail freedom remains explicit. ARCEP tells professional users that they may choose their current operator or another and need not accept the first migration proposal immediately (ARCEP's professional-user guide). A deadline creates urgency, but it does not erase choice. The compact asks commercial operators to move customers off the common copper network while preserving competition over the replacement.

That is harder than a captive conversion campaign, yet it is essential to making closure a regulated market transition rather than an asset owner's private cutover.

The deeper point is that wholesale access and public-service continuity are connected. A local authority or business does not experience the market as a regulatory diagram. It experiences whether a suitable offer can be obtained at a location, whether specialist equipment can attach to it, and whether the responsible supplier will support the required service. Wholesale maturity expands the set of credible retail answers.

When ARCEP requires both layers before closure, it turns competition policy into operational redundancy: more than one organisational route can be available to solve a migration problem.

Migration is a chain of hand-offs

ARCEP's allocation of roles reads less like a hierarchy than a relay. Orange owns the copper network, initiates closure and manages its operational steering. Fibre-infrastructure operators align replacement deployment with the plan. Commercial operators hold the direct customer relationship and organise migrations.

Local elected officials are close to residents and can help identify people who are difficult to reach or need particular support, although sector communications remain primarily the operators' responsibility (ARCEP on closure organisation; ARCEP's local-authority guide).

Each hand-off has a different failure mode. Orange can publish a lot whose local constraints have not been sufficiently surfaced. A fibre-infrastructure operator can leave an address difficult to connect or information about it unreliable. A commercial operator can fail to identify or engage a customer in time. A municipality can overlook a copper-supported public contract. A business can migrate broadband yet miss the line attached to an alarm or lift.

The operating compact cannot eliminate those risks with a single command because no single actor holds all the relevant knowledge.

That is why “customer migration” is too narrow if it means changing a broadband subscription. ARCEP advises enterprises to inventory every contract resting on copper: ADSL, VDSL or SDSL internet; analogue or digital telephony; a switchboard using the public switched telephone network; connected equipment such as lifts, surveillance systems, payment terminals and emergency lines; analogue lines used to contact fire and rescue services; and intersite links where applicable.

The current operator or operators are the primary interlocutors for that inventory and for replacement offers (ARCEP's professional-user guide). Migration is discovery before it is installation.

Data sharing provides the common operating picture. Under ARCEP's December 2023 market decisions, Orange must transmit information at defined lot stages to infrastructure operators, commercial operators, territorial authorities and the regulator. A public trajectory file records the data used to construct lots and tracks possible postponements. A file on non-connectable premises supports commune- and building-level oversight of remaining fibre work.

A correspondence file helps match copper addresses that lack a fibre solution and lets commercial operators request address corrections from infrastructure operators (ARCEP on closure obligations).

These are not administrative by-products. They reduce the information asymmetry created by Orange's ownership of the legacy network and by fragmented knowledge of the replacement. The trajectory file gives a local authority and competing operator a view of intended sequence. The non-connectable-premises file converts a general coverage claim into a list of exceptions that can be worked. Address correspondence prevents a copper location from disappearing between incompatible records.

Shared data turns accountability from “someone should know” into a question of who received which gap at which stage.

The framework adds a local escalation for the hardest migrations. Approaching technical closure, a commune can request information from Orange on its still-active copper base so it can assist more complex cases, including people distant from digital services or living with disabilities; ARCEP specifies that some of this information is available from 12 months before closure under a confidentiality agreement (ARCEP on closure obligations). The municipality does not thereby inherit the operator's migration duty.

It contributes local knowledge where a purely commercial contact process may be weakest.

This boundary matters ethically and operationally. Local government is often the most visible institution when residents are confused, yet ARCEP says information, support and migration actions belong to Orange and the commercial operators with whom users subscribe. Elected officials help because they know the territory, not because the network industry may transfer its obligations to them (ARCEP's local-authority guide).

A sound compact invites municipal participation without using proximity as an excuse to privatise success and socialise difficult cases.

Public-service continuity is where the abstraction ends

For a municipality, “copper” may not appear in an asset register as a strategic category. It is embedded in subscriptions and services acquired at different times by different departments. ARCEP warns that technical closure stops services based on ADSL, VDSL, SDSL and the legacy switched telephone network.

It lists buildings open to the public, schools, administrative services, videoprotection and teleassistance among the local uses that may be affected, and urges migration to fibre or another available technology as early as possible (ARCEP's local-authority guide).

The inventory problem is organisational before it is technical. Connectivity may have been purchased as part of a security service, lift-maintenance arrangement or facilities contract rather than as a recognisable telecoms line. Responsibility may be divided among an IT team, procurement office, school, building manager and external provider. A commune-level deadline gives those units a reason to compare records.

The local authority's task is not to reproduce the operator's network records; it is to identify the public functions for which loss of a hidden copper bearer would matter.

Public procurement adds temporal friction. ARCEP specifically advises vigilance when contracting for services that rely on copper, especially where the contract runs for a long period (ARCEP's local-authority guide). That warning turns closure dates into procurement constraints. A service contract cannot be judged only on whether it works on signing day; it must remain supportable through the scheduled retirement of its communications layer.

Migration may require a contract amendment, a replacement procurement, site works or coordination with a specialist vendor, all of which consume the notice window.

Local governance supplies forums for resolving such dependencies without pretending they are purely municipal. ARCEP describes a national fixed-networks consultation committee involving public authorities, fibre-infrastructure operators, commercial operators and associations of elected bodies. At territorial level, departmental consultation committees convened and chaired by prefects allow elected officials and operators to exchange information and follow operational implementation.

Orange's regional delegates also hold exchanges with elected officials, while departmental structures monitor the plan at a wider local scale (ARCEP's local-authority guide; ARCEP on closure organisation).

These forums are valuable because continuity questions rarely respect corporate boundaries. A school may depend on a municipal contract delivered over infrastructure operated by one entity and sold by another. A teleassistance service may involve an equipment provider as well as a connectivity supplier. An address may be shown differently in copper and fibre records. The committee cannot perform each migration, but it can force a problem into a shared field of view and identify the actor able to act.

Continuity also includes the old network during the interval. ARCEP's insistence that acceptable copper quality remain available for households and businesses still dependent on it prevents the future fibre system from becoming an alibi for neglecting present service.

At the same time, the regulator calls a well-built and well-operated fibre network a condition of successful substitution and notes commitments by operators to improve technician training, intervention controls, connection quality and degraded infrastructure (ARCEP's local-authority guide). The compact therefore spans both sides of cutover: maintain what is still needed, and make the replacement dependable enough to inherit critical uses.

No frozen evidence supports a claim that every public service moved without disruption. Nor is such a claim necessary to recognise the design's strength. The relevant test is whether dependencies were made discoverable, whether actors received usable notice, whether alternatives could be ordered, and whether a failed readiness test could change the date. Public-sector continuity is not a celebratory outcome asserted at national level. It is a continuing obligation recreated in each commune.

The terminal problem: lifecycle and lock-in beneath the line

The most stubborn copper dependencies can sit behind a working line. A private branch exchange may support a building's internal telephony but not speak the internet protocol used by replacement services. An alarm panel or lift communicator may have been installed on the assumption that an analogue bearer would persist. A payment terminal may be contractually tied to a service arrangement whose communications component is barely visible to the user.

These are lifecycle problems: the access network can be ready while the terminal or surrounding service is not.

ARCEP tells professional users with a non-IP-compatible PABX to prepare a migration plan with an integrator or operator before copper closure. For alarms, surveillance systems, payment terminals and lifts, it separates two requirements: migrate the telecoms access and ensure the terminal is compatible with the new technology. It notes that solutions have been developed for most special uses, but that does not turn compatibility into an automatic property of equipment already installed (ARCEP's professional-user guide).

This secondary lens reveals a form of lock-in that wholesale access rules alone cannot solve. Competition among fibre offers may be healthy, yet a user can remain constrained by an ageing switchboard, a maintenance contract or a terminal vendor. Conversely, replacing a terminal without arranging suitable connectivity and resilience leaves the migration incomplete. The service is a stack: access line, powered equipment, protocol compatibility, operator offer, specialist maintenance and organisational procedure.

Closure testing must follow the whole stack to the human outcome it supports.

Emergency communication makes the point sharply. ARCEP says certain establishments open to the public must be able to alert public fire and rescue services immediately, under requirements that vary by establishment category. Voice over IP on fibre or another network can satisfy the communication duty, but the establishment must ensure continuity of electricity for the terminal and box during a power cut, potentially through backed-up power, a battery or an external generator (ARCEP's professional-user guide).

Substituting an access technology changes the power assumptions around the service; continuity must be engineered, not inferred from bandwidth.

The same logic applies without dramatics to ordinary professional sites. ARCEP notes that a premise without electricity needs an electrical connection or a recharged battery system for fixed telephony to continue after copper. It also advises users facing environmental constraints—temperature, electromagnetic disturbance or limited space—to state them to the commercial operator so an adapted equipment arrangement can be proposed (ARCEP's professional-user guide). A national fibre map cannot answer either question.

They emerge only when the planned service meets the actual site.

This is why the closure programme should not be reduced to persuading reluctant users to adopt a newer product. Some users need basic telephony rather than an internet bundle; ARCEP says such offers exist, though each commercial operator decides whether to market one (ARCEP's local-authority guide). Others need a redesign of a specialist system. The compact must preserve the difference between preference, technical compatibility and legal continuity.

Treating all three as a sales conversion would obscure the very dependencies the notice period is intended to resolve.

For Orange's leadership, terminal lock-in sets a limit on central control. The company can establish lots, publish data and coordinate its own teams; it cannot unilaterally modernise every third-party device or public contract. What it can do is design the programme so those constraints surface early, give commercial operators and users a credible interval to resolve them, and accept regulatory postponement when the network-level prerequisites are missing.

Industrial execution here means standardising the discovery and escalation of non-standard cases.

Data, non-discrimination and postponement form the control loop

An operating compact needs feedback, not just assignments. France's closure framework obtains it through three linked controls: shared data exposes the current state; non-discrimination constrains how Orange uses its position; and postponement changes the schedule when the evidence fails. Together they prevent the plan from becoming either a voluntary coordination exercise or a date imposed without an alternative.

Transparency begins before officialisation. Orange must publish the trajectory file containing the data used to construct lots, including its criteria, and update the file as the programme evolves. Stakeholders can then challenge proposed membership during the sharing phase. After officialisation, the notice clock gives operators and users time to act.

Closer to closure, files on non-connectable premises and address correspondence focus attention on specific gaps, while information about active copper can support municipalities working with difficult migrations (ARCEP on closure obligations). The data becomes progressively more operational as the deadline approaches.

Non-discrimination is the rule that keeps the loop open to rival evidence and rival service. ARCEP supervises the principle during lot construction, and the regulatory criteria require the wholesale and retail availability needed for users across the closure area. Orange cannot define readiness solely through the parts of the fibre market it controls or through its own retail capacity (ARCEP on closure obligations).

In practical terms, an objection from another infrastructure or commercial operator can reveal a migration constraint that a vertically narrow view would miss.

The postponement duty gives that evidence force. If at least one closure criterion is not met at the announced time, Orange must move the closure and announce a new date. ARCEP can examine compliance throughout the process. The framework also limits deferral: after 24 months, closure may proceed with at least three months' notice once very-high-speed alternatives capable of providing replacement connectivity are available (ARCEP on closure obligations). The system therefore neither treats the first date as absolute nor grants an unbounded veto.

That balance is economically significant. An unconditional deadline would transfer the cost of incomplete readiness to users and competing operators. Unlimited postponement would leave Orange maintaining parallel networks without a credible exit and weaken the incentive to complete difficult fibre work. The regulated mechanism places the burden first on meeting common criteria, then sets a bounded route through persistent exceptions with alternative connectivity.

It is a controlled transition, not a guarantee that fibre will be the answer in every circumstance.

The first lots show why public reporting should distinguish corrections from failures. Three communes in the first lot moved to a later technical date; dozens in the second lot did not pass commercial closure on the initially intended date because the criteria could not be assured (ARCEP on closure organisation). Those decisions are evidence of incomplete readiness in particular places, but also of a process able to register it.

A programme that never postpones may be exceptionally prepared; it may also be suppressing inconvenient information. The surrounding data and criteria are what let observers tell the difference.

This control loop is the most consequential link between Heydemann's strategic commitment and local continuity. Leadership supplies the organisational insistence that the work reach an end. Regulation makes the route conditional on evidence external to the executive promise. Teams and operators act on the resulting exceptions. Local structures contribute knowledge that central systems may lack. No part alone is sufficient: ambition without controls is hazardous, while controls without an executable programme merely document delay.

Responsibility must remain where the work sits

Writing about a chief executive creates a temptation to make every operational outcome an extension of personality. The copper programme resists that treatment because its public architecture assigns duties with unusual clarity.

Heydemann's attributable role is to have presented a plan that included copper decommissioning in France by 2030, publicly framed the first plan year as one of collective mobilisation, and later described the full-fibre transition as an industrial decommissioning project (Orange's 2023 strategic-plan announcement; Orange's 2023 integrated-report interview; Orange's 2026 retrospective).

Those statements create a legitimate standard for judging executive direction. Did Orange treat retirement as a core transformation? Did it organise recurring lots, data, stakeholder engagement and internal coordination? Did its public language recognise the collective and industrial character of execution?

Orange's governance page makes the executive committee led by Heydemann responsible for the daily implementation of major strategic orientations and transformation projects, including operational-risk oversight and resource allocation (Orange's governance page). That locates programme-level accountability without pretending the page independently verifies a copper outcome.

The next level belongs to Orange teams and the other named actors. Orange manages closure of its asset and supplies mandated information. Commercial operators communicate with their subscribers and organise migrations. Fibre-infrastructure operators resolve deployment and address readiness. Local authorities inventory their services and contribute knowledge of territory and residents. Businesses, building managers and specialist suppliers deal with installed terminals and service procedures.

ARCEP defines and monitors the conditions, notice, data duties, non-discrimination and postponement (ARCEP on closure organisation; ARCEP on closure obligations).

Keeping those boundaries intact improves analysis rather than diminishing leadership. A chief executive's quality in a distributed transition lies partly in constructing an organisation that does not require personal intervention in each commune. The relevant achievement is not heroic proximity to every cutover; it is an operating design in which local knowledge travels upward, duties travel outward, and exceptions alter decisions before service disappears.

Conversely, a national plan cannot claim operational success merely because management repeated the destination. The burden of proof remains with the condition and outcome at the appropriate layer.

The boundaries also protect local actors from misplaced blame. Mayors are visible to residents, but ARCEP says operator communications and user migration remain responsibilities of Orange and commercial operators. Municipal knowledge is a valuable supplement, especially for people requiring particular support, not a replacement for the sector's duties (ARCEP's local-authority guide). Likewise, a commercial operator cannot solve a missing fibre address alone; the infrastructure operator and shared data process must participate.

This responsibility map is what turns “trust” from executive language into an operational test. Orange's 2023 plan announcement used trust as one of its guiding principles (Orange's 2023 strategic-plan announcement). In a shutdown, trust is not confidence that no problem will occur. It is confidence that the schedule is visible, the evidence can be challenged, a suitable replacement can be ordered, the old service will be maintained until the gate, and a missing criterion can trigger a different date.

What the 2023–2025 compact demonstrates

The execution evidence from 2023 to 2025 supports a bounded conclusion. At company level, copper decommissioning was embedded in the strategic plan and publicly treated as collective execution. At regulatory level, the December 2023 framework connected copper dates to complete fibre, wholesale and retail availability, notice, transparency and non-discrimination.

At operating level, the first annual lot moved through commercial and technical closure in 2024 and 2025 with several local technical dates deferred, while the next lot's 2025 commercial milestone excluded communes whose criteria could not be assured (Orange's 2023 strategic-plan announcement; ARCEP on closure obligations; ARCEP on closure organisation).

That is not proof of universal seamlessness, nor does it answer questions the frozen evidence does not measure. It does show the shape of an executable retirement regime. A national horizon changed incentives. The commune made accountability locally intelligible. Lots permitted repetition and learning. The sharing phase allowed proposed scope to change. Separate commercial and technical gates prevented new dependence while giving existing users time. Wholesale conditions protected choice and alternative-operator preparation.

Data sharing made gaps visible. Postponement prevented the first announced date from overruling readiness.

The compact also shows why infrastructure retirement is a governance problem disguised as an engineering one. The physical act of stopping a network is at the end of the causal chain. Before it come decisions about scope, information, market availability, contracts, terminals, electrical backup, public duties and support for users who are harder to reach. Fibre is the principal replacement, but “fibre present” is only one of several required answers.

The project becomes industrial when those questions are asked in a consistent order for every lot without erasing the specific answer at each address or institution.

Heydemann's place in that account is consequential but precise. She attached leadership authority to the destination and presented execution as a collective undertaking. The operating compact beneath that commitment was produced by the interaction of Orange's programme, ARCEP's rules, operator markets and local governance. Its credibility rests on that plurality. A shutdown affecting essential communications should not depend on one company's assertion that it is ready, even when that company owns the network.

Conclusion

France's copper retirement is best understood not as the last chapter of fibre deployment but as a controlled withdrawal of a shared dependency. Heydemann's strategic choice made the exit explicit and industrial. The harder work then passed through communes, annual lots, two closure gates, wholesale tests, migration inventories, local forums and continuity checks.

Orange teams, other operators, infrastructure owners, authorities and users each held information the others needed; ARCEP's notice, transparency and postponement rules gave that information the power to affect timing.

The compact's discipline is therefore visible in both movement and restraint. It can advance a national programme, yet stop a local date from outrunning the replacement. That is the operating achievement worth examining: not a claim that every migration was effortless, but a structure designed to discover what still depends on copper before the copper goes dark.

Sources