As experienced experts in IT implementations, as well as from our HR services business, we know the importance of proper data migration. You cannot perform your daily processes properly if the data is or has come into the system incorrectly. Colleagues Arjan Kats (left) and Denis Cojuhar (right) are part of our Data & Analytics team and are the right people to help you with a decisive data migration. 

What is a data migration?

“Data migration is the process of transferring data from one system or platform to another. This is often done when implementing new software or reorganizing IT infrastructure. Data migration is an important part of most payroll and HRIS implementations and is also relevant in the case of mergers and acquisitions. I myself often work on the technical side of data migration, where Arjan is active as the project leader or process lead,” says Denis.

Arjan adds: “The goal of any data migration is a smooth transition where the data is and must remain accurate, usable, verifiable and secure. To ensure this goes smoothly, we have identified several key aspects:

• Collaboration   

Often, HR implementations have a hard deadline and are under time pressure. Working through the project plan together is very important then. For example, the Payroll Stream provides the payroll code catalog, Systems Stream helps secure frozen data, and Test Stream takes test data.   

• Compliance  

Data is moved between different systems and sometimes stored, edited, and eventually deleted outside systems. The GDPR provides guidelines for this, but they must be applied to the specific circumstances of the project. Therefore, Privacy and Security are involved early on.  

• Quality Control  

The (Project) controller determines what requirements the evidence for proper data migration must meet. It also determines how long the migrated data must be kept, for example for auditing purposes. The controller is also involved at an early stage.  

•  Definition of “Done. 

When is the data migration ‘done’? Often, all data must be migrated correctly and completely one-to-one, but sometimes this is not feasible or efficient in reality. That is why the measuring rod that the data migration must meet is determined in advance. The same applies to the entry criteria: what must be met before data migration is started at all?  

• Data quality   

“Garbage in = Garbage out” certainly applies to data migrations. Before starting, a data quality assessment is done together with the data owner. Based on these results, a plan with the necessary interventions is drawn up.

As Consultday, we bring added value during data migrations by taking care of these 4 steps, among others:

1. Organizing the data migration
2. Setting up the processes
3. Finding specialist help, if needed
4. Stakeholder management

Migrating in phases

Arjan: “A data migration is divided into 3 phases, each with a number of steps that follow one another in a logical order. The three phases are

• Planning
• Organizing and
• Executing

Each phase has underlying steps and focal points, as you can see in our own framework. This framework helps you take a proper and structured approach to your data migration, from A to Z.”

Do’s and don’t’s with migrating 

Joint knowledge and talents provide a wide variety of tips. Arjan and Denis have the following do’s and don’t’s for you so that you properly map, organize, and execute your data migration. 

Do’s 

• Keep room for trial and error in the planning phase.   

Not everything works immediately as you expect or hope; you can learn from that. Create room for exploration and improvement.  

• Don’t be a data island.  

Find collaboration with other streams: payroll, testing, and the entire integration. Data migration affects the entire organization and is one of the critical success factors for overall integration.  

• Ask questions throughout the process.  

This do is related to the one above. Keep connecting with your (internal) end customers and users. Ask them all the questions you have and not just at the planning stage, do so at every step and stage.  

• Documentation can be learned.   

Don’t just document the data migration and what needs to be done for that. If necessary, also describe the definition of the migrated fields so that it is clear to what a piece of data means in the future.  

• Start early with data management.  

Very early (!), as soon as you know a project is coming up! This is especially true for increasing data quality.  

• Legacy systems.  

Get the technical resources and access to systems to extract data as soon as possible.  

• Use multiple cycles.  

Every data migration has its own challenges, especially if the target system is new. Therefore, test the data migration, preferably multiple times. And when testing, don’t forget to test that the data lands properly in connected systems too. 

Don’t’s 

• Don’t underestimate the effort.  

Extracting, mapping, loading, and validating data takes more time and effort than is often thought.  

• Don’t confuse wants with needs. 

Avoid the thought “everything must be migrated”. Focus on what is actually essential.  

• Don’t ignore data dependencies.  

These can completely disrupt loading processes. So, determine the right priority when loading: first the collective bargaining agreement, then the pay scale and step, then the pay codes. 

Starting your decisive data migration? 

Have Arjan and Denis made you curious about the possibilities that a decisive data migration can offer your organization? We would love to help you with this. The actual first step is to contact us, we are happy to start this conversation with you! 

---