If you are undertaking an ERP or major system transformation programme, unless you are a greenfield site it is likely that data migration is a core workstream of your initiative.
It is vital that data migration is considered right at the start of the project as it will have key input into the scope and shape of the programme. Consider it late and it is highly likely that there will be scope changes and delays to your go-live.
As part of your scoping phase, you need to determine your overall data migration strategy, and how it feeds into your cut over, contingency and go-live plans.
Considering your data migration strategy early will not only bring shape to the overall programme; it is a real opportunity to find additional benefits through early identification of scope.
Why look early?
Key to the data migration is the quality of the data and where it is located. There is a good chance that you have a lot of data outside of your core system, some of which may be in old legacy applications or in spreadsheets (possibly supporting long term work arounds, which were always intended to be short term!). Whilst the migration should not be a ‘catch all’ leading to a significant and complex scope, by looking early you will understand the data landscape and you can make conscious decisions on scope. You should also use this time to identify additional benefits, such as through decommissioning of old systems and spreadsheets.
Key considerations
What other primary areas should you think about early in your data migration strategy? Here are a few:
- How much data are you going to migrate? What regulation drives this decision? Will you migrate all historical data or, for instance, just the last financial year? If the latter, how will you access data older than one year? These are important decisions as they will determine the complexity, the cost and the risk profile of your migration, and may have operational impacts once live
- How good is the quality of the data? Do you have duplicates? Do you have work arounds due to the data quality? By assessing the data quality early, you can then plan how you will cleanse, enrich, de-duplicate or delete data prior to migration, making the migration go-live smoother, and significantly reducing the risk to the business at cut over
- Who owns the data? Do you have people that understand the details? Are these people on your project? Migrations are a blend of business knowledge and technical knowledge, and a team needs to be put in place to reflect this. Migration is not a pure technical exercise owned by IT, with the data being owned by the relevant Business area as part of the transition to BAU
- When are you going to migrate? Month end or Financial Year end? Or some other time? For a finance transformation programme, often the year end is picked so that year end balances are cut over and the new year is started on a clean slate. If you do not migrate at year end, the complexity of the migration will increase. These decisions will drive overall programme planning
- Does everything have to be migrated in one go? Can it be phased? Is it possible to leave some until after go-live? It may be possible to perform some migration ahead of the migration window (for instance master data such as vendors and customers) reducing the work during the cut over period. It is imperative that you do everything you can to de-risk the final go-live migration exercise
- How can you be sure my applications will work on the migrated data? This is key, and several dress rehearsals should be planned in through the project. A good approach is to build the migration up over time, using a small set of data first then expanding as you get confidence in the process, the data quality and the operation of the new target system. Try to find as many issues as you can on a smaller set of data, operating in an agile way.
A safe transition
Whilst there are many elements that feed into the overall data migration strategy, it is crucial that it is treated as a core workstream in your ERP or major system implementation. Think about it early; this will help to form a robust scope for your project, drive the overall planning and will feed into the overall risk profile, which can then be managed. We know that data migrations are typically not easy, but by doing them in the right way, with the right people, you will ensure a safe transition and successful go-live for your business.
Do you need any migration or transformation expertise?
- We are pragmatic. We understand the components and approaches needed for a workable solution, however compromises will always have to be made in your data migration strategy. We can bring our experience, working client-side to help assess the options and agree a pragmatic approach. Whilst there is a lot of science behind the data migration tasks, there is also an art to getting all stakeholders and various competing requirements balanced
- Data ownership is never an easy one to tackle; it is possible that there is no real ownership. Again, we can drive pragmatic, experienced led solutions to work through the ownership issues to support a successful data migration
- There will be a constant set of parallel activities in data migration with the technical ETL activities working alongside the business-led cleanse and validation activities. Weaving the two together is necessary and we can bridge the two to have one integrated approach. There are many specialist companies and System Integrators who can deliver data migration solutions, however typically these focus on the technical activities and tooling; the business side often gets omitted, which is something that Project One can deliver.
To discuss further, please contact Nick Houlton
Mobile: 07716 075616 | email: nick.houlton@projectone.com
