HM Revenue and Customs is preparing a support and change contract worth about £500 million around the National Insurance and PAYE System, committing substantial spending to a critical legacy platform while trying to make the surrounding tax architecture more adaptable.

The preliminary market engagement covers application support and maintenance alongside change capability for strategic milestones, legislation, and other business requirements. The proposed agreement is expected to be worth about £501 million including VAT, with an initial five-year term and an option that could extend it by a further two years.

Rather than replacing NPS outright, the contract is intended to sustain the existing estate while supporting continuing change around it. HMRC has described the platform as operating on legacy technology with limited modularity and scalability, while its longer-term direction is towards a more service-oriented architecture capable of supporting more responsive tax processing.

A supplier-engagement process is expected before the formal competition, with the resulting service due to begin in 2027. The procurement therefore sits in the difficult middle ground of public-sector modernisation: HMRC needs to change the architecture, but it cannot suspend tax and contribution administration while the work takes place.

Legacy systems still carry core services

NPS has been operating since 2009 and supports work across HMRC and the Department for Work and Pensions. Systems of that age can accumulate business rules, interfaces, customisations, and operational workarounds that make replacement more difficult than the underlying software architecture alone would suggest.

That constraint explains why large public-sector programmes can spend heavily on maintaining ageing systems while simultaneously pursuing modernisation. The existing platform still has to absorb legislative changes, security work, operational demand, and routine maintenance even while teams attempt to separate functions or introduce newer components around it.

Tax infrastructure makes the problem especially unforgiving because policy deadlines do not move simply because a technology programme does. Changes to rates, allowances, reporting, employer obligations, or calculations can require system changes on fixed dates, leaving little scope for the department to defer implementation while an architectural migration catches up.

The proposed NPS agreement reflects that reality by combining support with ongoing change rather than pretending the legacy platform can be frozen until a replacement arrives. The difficult question is whether those changes gradually reduce dependence on the monolithic estate or make it harder to retire by adding another generation of functionality.

Modularity has to be created while operating

A service-oriented architecture offers a route towards smaller and more replaceable components, allowing individual functions to evolve without every change running through the entire platform. In principle, well-defined interfaces can let newer services coexist with parts of NPS that cannot yet be retired.

However, technical modularity depends on decisions about business ownership as much as software design. HMRC has to determine where particular rules belong, which system remains authoritative for each piece of data, how services recover when dependencies fail, and how responsibilities are split between departmental teams and suppliers.

The contract structure will therefore deserve as much scrutiny as the target architecture. An agreement lasting up to seven years could provide continuity through a difficult transition, but it could also deepen dependence on the winning supplier unless HMRC retains control over documentation, interfaces, code, architecture, and operational knowledge.

For bidders, the opportunity is substantial but constrained. The successful provider will be expected to maintain a nationally important live service, implement policy against statutory deadlines, and support architectural evolution without treating modernisation as permission to destabilise the platform already processing tax and contribution records.

HMRC’s proposed spending does not buy a clean break with legacy technology, and portraying it as a £500 million replacement would overstate the procurement. Instead, the department is trying to create room for future architectural change while keeping the existing system operational, leaving the quality of that transition as the measure of whether the money reduces future dependency or simply supports it for longer.