Not all policy data should be migrated.
A view from Owen Greenwood, Client Partner – Insurance
When insurers start planning a policy administration system transformation, much of the early conversation naturally centres on the destination. Which platform are we moving to? How quickly can we retire the legacy estate? How will we get the data into the new environment?
Insurers need to ask those questions. But in our experience, they often ask them too early.
Before an insurer decides how to migrate its policy data, there is a more fundamental question to answer: what actually needs to move?
It sounds simple, but it is one of the most important decisions in a PAS transformation. An insurer operating multiple legacy platforms does not automatically need to move every policy, every piece of history and every dataset into its future-state system.
Some books clearly need to migrate. Insurers may choose to let others run down. Some data may still be strategically important, but it may need remediation before a sensible migration can take place.
The mistake is assuming migration is the default answer simply because the data exists.
Our view is that PAS transformation should begin with a strategy for the data estate, not an instruction to empty the legacy platforms.
In our first perspective on PAS transformation, we explored why policy administration systems can become where transformation programmes go to die when the complexity of insurance data is underestimated. The next challenge is more specific: once you understand what is in the estate, how do you decide what deserves to become part of the future one?
Migration is not one decision
A multi-PAS estate can contain very different situations.
One system may support a strategically important book that clearly needs to move. Another may support policies that are nearing the end of their lifecycle. A third may hold data where quality issues are significant enough that the organisation first needs to decide whether remediation is justified.
Programme teams may find it convenient to treat all three as identical migration problems, but that approach can create unnecessary cost and complexity.
The better approach is to make an explicit decision for each relevant book, system or dataset.
Broadly, that decision falls into three areas:
- Migrate — the data belongs in the future state and is fit to move.
- Assess further — the organisation may need the data in the future state, but it needs to understand its quality or complexity before deciding the right migration path.
- Run off — migration would create more complexity than value, so the organisation keeps the book in its existing environment for a defined period.
Once the organisation knows what needs to migrate, there is a second question: in what order should those migrations happen?
That is where sequencing becomes a strategic programme decision.
Migrate when the data belongs in the future state
Some migration decisions are relatively clear. A book remains important to the business, the insurer still needs to service active policies, and the future operating model requires the relevant data.
In that situation, migration is likely to be the right answer.
But even then, “migrate the book” does not necessarily mean “move everything that exists in the source system”.
When insurers move towards platforms such as Acturis, Open GI or Insly, the target data model may differ materially from the legacy source. The objective should therefore be to understand what the future platform needs and how the source data supports it, rather than trying to find a new home for every field simply because it exists today.
This is an important distinction. A successful PAS migration should create the data estate the insurer needs next, not recreate the legacy estate as faithfully as possible.
That means understanding both sides of the migration: the structure and quality of the source data, and the expectations of the target model.
Only then can the organisation make an informed decision about what needs to move and how it should be transformed along the way.
Assess further: when data quality changes the decision
Some of the harder decisions arise when teams struggle to understand the data itself or find poor-quality records.
In a multi-PAS estate, one source may be relatively straightforward to migrate while another may contain quality issues significant enough to make the organisation question whether it should migrate the data at all.
That is why data quality needs to be understood before migration scope is fixed. The first question is not automatically how to clean everything that exists, but what the condition of the data means for the migration decision.
Some data may need further assessment before the organisation can agree a migration approach. Other data may present a clearer path.
The important point is that the organisation makes that decision deliberately before attempting to migrate everything.
Run off: when NOT migrating is a better decision
Transformation programmes sometimes treat run-off as the lesser option, as though successfully moving everything is the ultimate measure of success.
We do not think that is the right way to look at it. For the right book, an insurer can make a deliberate and sensible strategic choice to allow policies to run down.
For some books, migration may not be necessary at all. Across a larger PAS estate, insurers may instead choose to let certain systems or books run down.
That does not remove the need for control. The insurer still needs to understand, manage and access the data appropriately while the book remains in place.
But it does mean PAS transformation doesn’t need to force every part of the estate down the same path.
A PAS programme should not be judged by the percentage of legacy data it manages to move. It should be judged by whether it leaves the insurer with the right estate for the future.
That is a different measure of success
Decide what moves first
Once the organisation has decided which books and datasets need to migrate, sequencing becomes important.
Where several PAS platforms are involved, the organisation should make the order of migration part of the overall strategy rather than decide it one project at a time.
The right sequence will depend on the particular estate and future-state strategy. What matters is that the organisation understands the wider migration landscape before deciding when each system or book should move.
As the programme progresses, the team should also build on what it has already learned about the target platform, the source data and the insurer’s own estate.
Don’t confuse moving more data with making progress
Large transformation programmes naturally create pressure to demonstrate progress. Leaders can see systems retired, records moved and milestones completed. But PAS migration needs a more considered definition of progress.
Moving every available piece of legacy data does not necessarily create a better future estate. In some situations, it simply transfers years of complexity into a new platform.
The more useful question is whether the organisation understands what it needs to carry forward and why.
That may mean making decisions that initially feel harder than simply putting everything into scope. Which books genuinely justify migration? Where does remediation create enough value to warrant the investment? Where would run-off be more sensible? And among the migrations that remain, which sequence gives the programme the strongest path forward?
Those are not merely technical migration questions. They are decisions about the future insurance business and its data.
That is exactly why they should happen early.
The real work starts BEFORE migration
Moving the data itself may be the most visible part of a PAS transformation, but teams create some of the most valuable work before the first migration run.
That means understanding the estate, assessing the data, deciding what belongs in the future state, and identifying where remediation is justified or run-off is the better answer.
Then, for everything that needs to move, deciding the right sequence. Together, these decisions help migration become more controlled and purposeful.
They also stop the organisation from investing time in solving problems that do not need to be carried forward.
A new platform does not automatically create a better data estate. The decisions made before migration do.
That is why the most valuable question at the start of a PAS transformation may not be:
How quickly can we migrate this?
It may be:
Should we migrate it at all?
Because the goal should not be to move the maximum amount of legacy data. It should be to carry forward the right data, in the right way, for the insurance business the organisation wants to become.
How Dufrain can help
From early-stage PAS strategy and platform selection through to data migration and legacy retirement, Dufrain helps insurers make better decisions, reduce delivery risk and create a more repeatable route to PAS transformation.
Where multiple migrations are moving into the same target platform, that repeatable approach can create significant efficiencies — in the right circumstances, we believe around 70% efficiency across subsequent migrations is achievable.
Talk to us about your PAS transformation plans, or ask us how a repeatable migration approach could help reduce risk, accelerate legacy retirement and improve delivery efficiency.
Read Owen’s previous perspective on PAS transformation