Summary

• HMRC says it is marginally ahead of plan on moving citizen interactions towards digital channels.
• Reaching 90% digital engagement depends on contact-centre-as-a-service and enterprise CRM platforms.
• The programme shows how public-service digital transformation increasingly depends on shared customer data, identity, support, and assisted digital design.

HM Revenue & Customs is slightly ahead of schedule on its plan to shift citizen interactions online, but the next phase will depend on major contact-centre and customer-management platforms rather than incremental website improvements.

The department’s proportion of digital interactions has risen to just under 80%, following investment intended to move the figure from about 70% to 90% by the end of the decade. Reaching the next stage will require more than persuading users to choose an online route. HMRC needs service channels, identity, case management, and customer data to work together across a complex tax system.

The department’s transformation roadmap sets out a digital-first service model for individuals, small businesses, tax agents, and intermediaries. It includes expanded online PAYE services, digital expense claims, inheritance tax modernisation, payment improvements, AI-powered digital assistants, GOV.UK One Login migration, and more personalised support through a new customer relationship management system.

The platform dependency is now becoming clearer. Contact-centre-as-a-service and enterprise CRM are expected to be fundamental to HMRC’s attempt to reach 90% self-service, giving the department reusable platforms across the organisation and reducing the need to ask customers for the same information repeatedly.

That ambition sits inside one of the UK’s most demanding public-service environments. Tax services are high-volume, rules-heavy, and often stressful for users. A digital-first model can reduce administrative burden when the journey is simple, personalised, and trusted. It can increase frustration when users cannot resolve edge cases, obtain reassurance, or understand why a decision has been made.

HMRC’s roadmap acknowledges that many customers are willing to engage digitally, but not every need can be met through self-service. It promises targeted support where digital self-service is inappropriate, including for vulnerable customers and people with more complex enquiries. That caveat is essential. A 90% digital target cannot be achieved responsibly by making phone routes harder to use.

The contact-centre element should therefore be treated as part of the digital operating model rather than a legacy channel. Voice, webchat, assisted digital support, case management, identity verification, and CRM data need to work together so citizens do not have to restart every time they move between channels. The quality of that orchestration will determine whether channel shift feels like service improvement or demand suppression.

Customer data brings a parallel governance challenge. A more rounded view of the user can help HMRC tailor support, reduce duplication, and identify where people need extra help. It also increases the importance of accuracy, access controls, transparency, and redress. Tax data is highly sensitive, and a centralised customer view has to be designed for trust as well as efficiency.

AI will add another layer to the programme, but it should not obscure the underlying transformation work. HMRC’s roadmap points to AI-powered assistants, adviser tools, voice analytics, robotics for triaging post, and synthetic data environments for innovation. Those tools may improve navigation and reduce administrative load, although they will only work well if core records, case workflows, identity systems, and escalation routes are coherent.

The programme illustrates a broader pattern in UK govtech. The first phase of digital government often meant moving forms and guidance online. The next phase is harder because it requires joined-up data, channels, identity, service design, and support around actual user needs. HMRC’s progress towards 80% digital interaction is meaningful, but the last stretch to 90% will reveal whether the department has redesigned the service or merely changed the front door.