Hong Kong is embarking on one of the toughest technology transformation initiatives since the introduction of the Octopus card back in 1997. The Special Administrative Region’s eMPF platform aims to reshape the administrative models of MPF (Mandatory Provident Fund) schemes and to standardize, streamline and automate the existing scheme administrative processes.
With a current membership of over 4.5 million people, and in excess of 300,000 employers using schemes managed by 13 independent Trustees, the transition to the eMPF platform and the associated transformation of scheme operating models will be no mean feat. The current plan is to complete the transition by 2025 and there will be a lot of moving parts and challenges to face as operating models are reengineered.
This article focuses on the challenge that Trustees face in wrangling and rearchitecting data to operate in a hybrid model, that is, both on and off premises. Whilst the canonical data elements used today may not change, the business processes, customer engagement model, and the underpinning data access and management functions, will all need to be revisited or re-architected. The transition will mean building an enterprise view of data, aligning to common eMPF data standards, conducting a potentially risky data migration cutover, and putting in place data delivery contracts and automation strategies.
At present, many organizations do not make use of enterprise conceptual models, let alone enterprise data models, but the transition to the eMPF platform makes this a good time to start.
This is because the migration and integration to the eMPF platform will change how data is produced and consumed across the entire organization and by the external pool of Trustees.
Designing an efficient operating model for the future will mean identifying new flows of shareable data across functional and organizational boundaries. Legacy integrations and system data interdependencies will need to be reviewed to gauge the potential impact of the changes, and whichever data repository was once considered the ‘single version of truth’ may no longer fulfil that role.
To scope the transformation required, organizations will need to apply enterprise conceptual modelling from both a system and data perspective, to both current and future eMPF states. Trustees will also need to revisit their enterprise data model to establish the impact of change to their information requirements for core business systems. A properly prepared enterprise data model may prove critical in identifying potential break points when aligning to the eMPF’s common data standard.
Data standards govern how data is managed, used, represented, formatted, defined, transmitted, structured and tagged. There are usually two key steps to imposing a common data standard.
The first is agreeing the standard, which often requires extensive business engagement and collaboration to obtain consensus, particularly when the data resides in multiple core business systems. The second is transforming and aligning data to comply with the common standard across core systems.
In the case of the eMPF transition, 13 independent Trustees are about to embark on this complex journey, with the saving grace being that step 1, the definition of the standard, will be nominated by eMPF. However, upstream and downstream systems will need to be modified to accommodate the new data standard, and the second step of data alignment to the common data standard will need to be carried out prior to the migration and cutover date.
For any organization, this kind of data alignment is a massive undertaking and can go horribly wrong if not managed properly. The transformation of customer data also needs to consider regulatory and compliance requirements in relation to personal data.
To establish the order of magnitude of effort required for eMPF data standard alignment, Trustee’s should consider profiling their existing data and conducting a comparative analysis to ascertain the variances as early as possible.
Migrating business-critical data is always potentially risky and, even today, data migration failures are not uncommon. In most of these cases, organizations have underestimated the effort and preparation required to mitigate the potential risks of failure.
Data migrations cannot be compared to dragging and dropping a file from one folder to the next on your desktop. They require a comprehensive strategy, a detailed plan and highly skilled people with a long history of execution.
In the case of the eMPF migration, organizations must invest time in getting to know their data, intimately understanding the target schema, rehearsing the migration multiple times, and conducting an audit and reconciliation after each rehearsal. The rehearsal and audits will highlight the unknowns and help set out ways of dealing with exceptions as they are uncovered – with the aim of eliminating surprises when conducting the final cutover.
After completing the initial data migration to eMPF, Trustee’s will continue to interact with the centrally hosted customer data, making interoperability critical. Data delivery contracts are one of the most important aspects of designing an efficient hybrid data architecture. At the time of writing, there is no indication of how this will be handled.
In principle, we recommend documenting contracts for all interfaces including data schemas (message formats and data) and transportation types, and their relationship to applications. This reduces the risk of issues arising from versioning and compatibility, while the fulfilment of each contract can be governed by some form of data handshake service.
The benefits of moving to the eMPF platform include standardizing and automating scheme administrative processes to reduce paperwork, improve data accuracy and transaction processing timeframes. Trustees will need to make major transformational changes to their target operating model (TOM), and workflow automation will become both an important enabler and a significant business opportunity.
One of the key success factors to any process automation is understanding the information lifecycle that underpins it, to ensure that the true benefits can be realized. The transition from manual to automated processing requires consideration not only of the workflow, but of the mode in which data is acquired, validated and transformed to comply with the defined data standards. However efficient the automated process, the old saying “garbage in, garbage out” continues to be true.
The eMPF timeline is well underway, and all Trustees and MPF administrators should now be defining their data migration strategy and plans.
We believe that top talent with relevant skills will quickly become scarce in the local market, as all 13 Trustees vie for the same resources. It is therefore important that Trustees act now to secure the right team to engineer a successful transition and to make sure they can reap the business benefits as soon as possible.