Case Study · 2025

Administration Transition for a Major Industry Superannuation Fund

Back to Newsroom

A major industry superannuation fund — serving approximately one million members across the health and community services sector — made the strategic decision to change its fund administrator. The transition involved migrating the full member registry, historical documents, and in-flight work items from the incumbent administrator's platform to the incoming administrator's target registry, effective 19 April 2025.

Running concurrently, an associated faith-based sub-fund of approximately 13,000 members was consolidated into the main fund as part of the same cutover window — adding a further layer of complexity to an already large-scale program. Desda was engaged across all three critical dimensions of the program: end-to-end ETL data migration delivery, independent reconciliation quality assurance, and program management across the full trial load cycle through to go-live.

~1M+
Members migrated
2
Funds consolidated simultaneously
34
Registry entity types migrated
3,908
Automated data quality tests

The Challenge

At approximately one million members, this was one of the largest administration transitions in the Australian superannuation sector. The program required migrating the full member registry — spanning accumulation, transition-to-retirement, retirement income, and pension products — together with more than 8.4 million documents (outbound correspondence, inbound member documents, images, emails, and in-flight work cases).

The incumbent administrator held member data across 416 source tables, of which 53 were mapped to the target platform. Data quality gaps in legacy records, limited control totals from the incumbent for some entity types, and the need to reconcile insurance cover at both a fund level and a member level created significant complexity. All of this needed to be resolved under a compressed timeline, with a fixed go-live window over the ANZAC Day long weekend.

Migration Scope
  • ~1 million accumulation, TTR, retirement income, and pension members
  • ~13,000 members from an associated faith-based sub-fund — consolidated simultaneously
  • 8.4M+ documents across outbound correspondence, inbound records, images, and WIP cases
  • 34 registry entity types across member, employer, adviser, insurance, investment, and document data

Key Complexities

  • Dual-fund consolidation in a single cutover window. The sub-fund consolidation required member deduplication across registries, reindexing of linked documents, and a separate insurance treatment — members of the sub-fund were transitioning to a new insurance policy at go-live, not simply carrying their existing cover across.
  • Incremental pre-load architecture. Rather than a single big-bang load, the program used an incremental pre-load approach: data was loaded progressively across multiple data drops, with only changes ingested in the final cutover window. This reduced cutover risk but required a more sophisticated ETL pipeline with robust change-detection logic.
  • Multi-layer medallion ETL pipeline. Data moved through Raw, Bronze, Silver, and Gold layers using Databricks, with over 3,900 automated quality tests embedded across each transformation stage — from null checks and format validations through to complex regex pattern matching.
  • Reconciliation gaps with the incumbent. The incumbent administrator did not provide control totals for all data entities in scope. Desda managed a structured gap resolution process — obtaining bespoke reconciliation logic, agreeing exceptions with all parties, and escalating residual gaps via a formal risk register.
  • Financial accuracy target of 99.999%. The program set a zero-tolerance standard for unexplained financial variances at go-live, requiring full traceability through every transformation layer — from source extract through to the final load into the target platform.
  • ANZAC Day weekend go-live. The cutover window fell over the ANZAC Day public holiday weekend, requiring detailed runsheet planning, pre-positioned teams, and clear go/no-go decision criteria based on reconciliation outcomes from the dress rehearsal.

Our Approach

Desda structured the engagement across three integrated workstreams: ETL design and delivery, independent reconciliation QA, and program management of the trial load and cutover cycle. Each workstream was interdependent — quality gates from the reconciliation function informed trial exit criteria, which in turn drove program sequencing decisions.

1. Strategy & Scoping
Established the full migration scope, including all 34 entity types, data drops, and trial load sequencing. Produced the ETL and reconciliation strategy documents defining the medallion architecture, control total framework, and DQ test approach. Agreed inclusions, exclusions, and known gaps with all parties.
2. Data Analysis & Mapping
Analysed 416 source tables from the incumbent registry, identifying 53 tables relevant to the target common data model. Produced source-to-target mapping artefacts for all 34 entity types. Identified reconciliation gaps where the incumbent could not provide control totals and initiated resolution actions.
3. ETL Development & Trial Load 1
Built the Databricks-based ETL pipeline across Raw, Bronze, Silver, and Gold transformation layers. Embedded 3,908 automated DQ tests — covering accuracy, completeness, consistency, uniqueness, and validity — with automated dashboards reporting daily pass/fail/warn status. Trial Load 1 validated the sequential load of four entity types across a 10,000-member subset.
4. Trial Loads 2–4 (Scale Testing)
Trial Load 2 established the baseline quality benchmark across all members and entity types. Each successive trial improved reconciliation coverage, reduced data quality issues, and refined treatment plans for known exceptions. Trial Load 3 validated the incremental load architecture end-to-end. Trial Load 4 targeted zero unexplained variances and no P1/P2 issues — the threshold for dress rehearsal readiness.
5. Dress Rehearsal
A full simulation of the go-live process using production-equivalent data, confirming runsheet timings, reconciliation outputs, and go/no-go criteria. All reconciliation reports, exclusion logs, and sign-off artefacts were produced and approved by the relevant business owners across ETL, operations, finance, and insurance.
6. Go-Live & Hyper-Care
Production cutover executed across the ANZAC Day weekend on 19 April 2025. The incremental pre-load was applied, final reconciliations confirmed, and go-live sign-off obtained from all parties. Hyper-care support followed to address any post-migration data queries and to hand over the reconciliation framework to the incoming administrator's operations team.

Outcomes

The program went live as planned on 19 April 2025 with no P1 or P2 issues at cutover. All ~1 million main fund members and ~13,000 sub-fund members were successfully migrated to the target platform, with the sub-fund consolidated into the main fund registry in the same window. All 34 registry entity types were fully reconciled, with financial accuracy at 99.999% and zero unexplained variances recorded.

The 3,908 automated DQ tests across the medallion architecture recorded a 0% hard failure rate at go-live — all issues identified during the trial load cycle were treated, agreed as known exceptions, or resolved prior to the production cutover. The incremental pre-load approach significantly reduced the cutover window duration, allowing the team to maintain service continuity through the public holiday period.

Go-Live: 19 April 2025 — Zero Critical Issues
  • ~1M+ members live on the target platform from Day 1
  • Sub-fund consolidation completed in the same cutover window
  • Financial data accuracy at 99.999% with no unexplained variances
  • 3,908 automated DQ tests — 0% hard failures at production go-live
  • 34 entity types fully reconciled and signed off across ETL, operations, finance, and insurance

What Was Achieved

  • One of the largest administration transitions in the Australian superannuation sector — ~1 million members migrated with zero critical issues.
  • Dual-fund consolidation delivered simultaneously, including a new insurance policy treatment for sub-fund members at transition.
  • A fully automated, multi-layer DQ framework with 3,908 tests providing continuous quality assurance across every transformation stage.
  • A structured reconciliation program that resolved gaps with the incumbent administrator and secured sign-off across all regulated financial and non-financial entity types.
  • Successful ANZAC Day weekend go-live, underpinned by rigorous dress rehearsal discipline and pre-agreed go/no-go criteria.
  • Reusable ETL pipeline, reconciliation artefacts, and treatment plan framework handed over to the incoming administrator's operations team.

"The structured approach to reconciliation — with clear metrics, gap resolution discipline, and daily quality dashboards — gave all parties the confidence to proceed to go-live on schedule."

— Client feedback

Why Desda

Desda's breadth across data migration, independent QA, and program management made us uniquely placed to lead this engagement. Our team brought hands-on experience with large-scale ETLR programs, deep superannuation domain knowledge, and the governance rigour required to coordinate across a fund trustee, an incumbent administrator, and an incoming administrator simultaneously.

  • End-to-end ETL design, build, and delivery capability
  • Independent reconciliation QA across 34 entity types
  • Program management across the full trial load and cutover lifecycle
  • Proven track record on Australia's most complex fund administration transitions

Planning a fund administration transition?

Get in touch with the Desda team to discuss your program — from ETL strategy and reconciliation through to go-live.

Talk to a partner