Summary

  • Mauritius’s draft would give one independent provider an exclusive licence to run the central number-portability system. It couples that structure with ownership limits, equal operator access, a long licence term and a proposed All Call Query design.
  • The draft names reliability, recovery, security, support and reporting obligations, but leaves applicants to propose much of the operating detail. The final call and award need measurable service levels and a public way to assign responsibility when porting or routing breaks.

A number that must travel

For a mobile customer, changing provider can be a straightforward commercial decision until the telephone number becomes part of work, banking, family contacts and public identity. Mauritius’s Information and Communication Technologies Authority (ICTA) says subscribers who switch today must change their number. Its draft licence frames Mobile Number Portability (MNP) as a way to remove that friction and strengthen competition.

The reform has moved beyond an idea. ICTA’s public MNP page says the licence category was created through 2026 regulations and that the Authority intends to award one licence to one successful applicant. The call for applications describes the selected provider as the builder and operator of a central system, on an exclusive basis. It also leaves the application deadline unfilled, and is still referred to as a draft on the Authority’s workshop page. No award, live service or operating results are established by the published material.

That distinction matters. The policy promise is consumer choice. The immediate procurement decision is about the control point that makes a transferred number reachable after a customer leaves one mobile network and joins another.

The control point is part of the call path

The draft specifies All Call Query (ACQ). Under the proposed approach, networks consult portability information to determine the recipient network for a number. The MNP service provider would establish a central clearing house and Number Portability Database, make current porting information available to relevant mobile, fixed and international operators, and explain how call scenarios would be routed. The application must also address person-to-person and application-to-person messages, SMS, USSD, international roaming and one-time passwords used by digital financial services.

Those requirements make the portability database more than a ledger of who owns a number. It becomes an input to routing and customer authentication flows. A port request can be accepted while downstream systems still hold stale information; a database can be available while an interface or synchronisation path is delayed. The draft itself asks applicants to explain the time between a donor operator’s acceptance, the central database update and the propagation of new data to operators that keep local copies.

Centralisation creates a useful common reference, but it also concentrates operational dependence. Replicated local copies can reduce reliance on a live central query, while increasing the need to prove that each copy is current. This is an engineering trade-off, not evidence that either model has been selected: ICTA’s October workshop agenda explicitly put central versus local databases, synchronisation, interfaces, data residency and scalability up for discussion.

The downstream surface is broad. ICTA’s draft asks for reliable routing of calls and messages, including mobile banking notifications and OTP delivery, and for service across Mauritius and its outer islands. It requests bulk porting for enterprises, query-performance indicators, disaster-recovery sites, redundancy, incident response, business continuity and a 24/7 support service. These clauses signal where a bad transition could matter; they do not prove that any such failure has happened.

Independence has to work in operation

The proposed safeguards begin with ownership. No public operator may own equity in the MNP service provider, directly or indirectly; the draft also bars the provider from holding equity in a public operator. The licensee must give all mobile operators transparent, non-discriminatory access. ICTA asks applicants to disclose shareholders and ultimate beneficial owners, demonstrate five years of operating portability systems, and submit a ten-year financial plan.

This separation responds to a real incentive problem: the operator that hosts the portability control point should not be able to favour its own retail network. Yet a no-equity rule alone cannot show that interfaces are equally usable, updates arrive on time, faults are handled consistently, or a fee structure does not burden one competitor more than another. Independence is a property of daily operation as much as corporate ownership.

The proposed term raises the stakes. The draft describes a 15-year licence, possible renewal for a further 10 years, and a deadline of nine months after award to make the service available. It asks for a cost model, planned charges and a financial forecast, but it does not yet publish a final charging rule or a numeric limit for query delay, availability, synchronisation or recovery. It calls for “high availability” and for applicants to provide reliability indicators such as failure rate and mean time to repair; much of the threshold design is left to the proposal and later regulatory requirements.

ICTA says applications will be evaluated through a comparative, qualitative “beauty contest”. Experience, neutrality, technical performance, resilience, implementation risk and financial sustainability all matter; the lowest-cost offer need not win. That may help avoid choosing a fragile low bid. It also makes the evaluation record important. If the Authority is selecting a sole provider for a long, exclusive term, applicants and operators need to understand how trade-offs are scored and how the selected provider will be monitored after the contest ends.

The public draft leaves choices visible

One clause requires locally hosted, on-premises infrastructure, then says a cloud-based solution is being considered following consultation. The contradiction is candidly visible in the draft, and ICTA’s workshop agenda lists data residency and infrastructure model as discussion topics. It would be premature to describe the final platform as local or cloud-hosted.

The question is not solved by a location label. Local hosting can support jurisdictional control while still depending on one site, one power supply or one operator. A cloud design can distribute recovery capacity but needs clear rules for data access, subcontractors, portability of records and supervisory inspection. The final licence should translate the selected design into obligations that can be audited, not treat “on premises” or “cloud” as a resilience result by itself.

Seychelles offers a bounded regional comparison, not a ready-made answer. Its 2023 portability regulations require a licensed clearinghouse to contract directly with operators, set cost sharing through business rules and retain porting records for at least seven years. They also allow the clearinghouse infrastructure to be located in or outside Seychelles. Those provisions show that rules can assign recordkeeping, commercial relationships and infrastructure location explicitly. They do not establish how well the system performs, and they should not be mistaken for Mauritius’s proposed terms.

Publish the accountability chain with the licence

The final call should turn the draft’s categories into verifiable commitments. For the routing layer, that means published thresholds for query latency, availability, database synchronisation, port completion, incident acknowledgement and recovery. For customers, it means a clear path to challenge a rejected or delayed request, recover service after an error, and know which party owns each step. For operators, it means equal interface conditions, version-change notice, auditable fee treatment and a dispute route that does not depend on the provider being challenged.

The regulator also needs an evidence trail after award: timestamped status changes, incident records, sample-based routing tests, recovery exercises, reasons for denials and periodic service statistics. Public reporting can protect sensitive network details while still showing whether the service meets its commitments. A launch date alone would prove readiness to begin, not reliable operation at scale.

This design has second-order consequences. A credible, predictable porting service can make customer choice easier to exercise and may increase competitive pressure. But a single exclusive provider also becomes a shared dependency. Weak cost recovery could undermine its financial sustainability; opaque fees could disadvantage operators; weak recovery terms could transfer outages to customers and other networks. None of those outcomes is shown today. They are the risks the licensing framework should make observable before the platform becomes indispensable.

The October workshop’s agenda is unusually concrete: architecture, synchronisation, OTP delivery, call routing, cost recovery, subscriber authentication, disputes, service levels and disaster recovery all have dedicated sessions. The public record reviewed here does not include the workshop’s minutes or a decision-by-decision response. The next public document should therefore do more than announce that stakeholders met. It should show which draft clauses changed, who carries each obligation and how compliance will be measured.

The central question is not whether Mauritius should make numbers portable. The draft has already moved toward that policy. It is whether a neutral provider’s exclusive control over the portability system will come with a clear, reviewable chain of responsibility. A number should remain reachable because the database, operators and customer process agree on its destination. When they do not, the licence must make it possible to identify the failure, restore service and hold the responsible party to account.

Sources