Data Migration in the Higher Education Sector (Part 2)

In Part 1, we ended with three common reasons why data migration projects in higher education often fall short: starting too late, overlooking reconciliation, and following unclear or inconsistent processes. These aren’t just technical missteps, but they point to a broader need for better planning, communication, and leadership. We’ll take a closer look at each of these challenges in this part.

Furthermore, we will focus on how Durham University successfully delivered a complex ERP data migration. Their approach highlights the importance of having the right people in place, establishing a clear and repeatable process, and investing time in thorough planning. For any university planning a similar transformation, Durham’s story offers practical, real-world lessons worth noting.

Starting Too Late!

There are so many moving parts in a university data project that it's no surprise that some elements are de-prioritised. But pushing the migration down the importance list is a false economy that all project managers realise at some stage.

As Steve pointed out, there comes a moment when the size and complexity of what is required become apparent. At the same time, the risk of not getting it right also comes into view!

As project deadlines approach, migration moves up the priority list, but at that stage, the challenge can begin to feel unmanageable.

The impact of not focusing on data migration until a late stage can be huge. The accuracy of the migration must be maintained, which means that more resources are required, or project deadlines need to be moved to allow for this stage to be done properly.

Recommendation:

Ensure data migration is a core part of any data project from the start. Engage with specialists and ensure there is a strategy in place to guide the entire process from the beginning.

No Time for Data Reconciliation.

No one is sitting at their desk wishing they had more work to do. Everyone is busy, and good people tend to be even busier.

So, when data migration projects are running through their cycles of testing, it’s a natural human instinct to focus on other more pressing things. This means that the data reconciliation may be rushed, not carried out thoroughly, delayed or, in some cases, not carried out at all.

This is far from perfect for the success of a project, yet it happens on even the best-run projects.

But if data reconciliation isn’t focused on properly throughout the project, problems go unchecked, issues go unfixed, and problems that could be identified get missed. This can lead to severe delays to projects, or at best gets picked up in hyper-care after the cutover, at that stage it’s an emergency firefighting problem and potentially a very costly one.

Recommendation:

Data reconciliation can be planned and scheduled to allow the relevant teams to plan their workload. Give clarity and ownership to both technical and business reconciliation; Don’t skip this stage and make sure you hit the deadlines.

Poor Process.

While large-scale data projects within universities need a diverse team involved, it’s important that roles and responsibilities are shaped early on.

For data migration, as Steve points out, the team leading that process needs autonomy and accountability to lead that process.

A good process should cover technical requirements, such as ETL, and human factors, too. As outlined above, ensuring the people who will participate in the process are considered from the off is a critical success factor.

Recommendation:

Always ask for the data migration team (or company) to outline their process, how they manage issues and examples of where it’s been used successfully before.

People, Process, Planning:

What the problems highlighted above show is that data migration is a people, process and planning problem.

People:

The people in your university and the contractors you work with will make or break the data migration project.

The best process in the world will fall if the teams responsible for data reconciliation aren’t given enough time to properly check the work.

Badly run meetings will ruin an implementation timeline. Poorly structured teams will slow down delivery. Poor communication can cause a project to grind to a halt.

These are all people, not data or process problems.

So, it’s essential to have a full understanding of who will be involved, what else they are involved with and how they can commit the relevant time to the project.

It’s also important to consider what outside expertise you will need. Picking the right partners who understand the sector, recognise the particular challenges of the HE sector and understand the rhythm of universities can help avoid numerous problems.

Process:

Data migration is often thought of as ETL - extract, transform and load. But when working with universities, the focus has to be much wider than just ETL.

It’s like moving house, but only focusing on putting the boxes in the removal van. Of course, your things will get to your new house, but without proper planning, packaging, labelling and unloading, the first six months in your new home will be a nightmare while you try and find everything.

ETL is part of the process, but not all of it.

It’s clear that for data migration projects to be a success, they need to focus on more than just ETL. The question of where and when to start a data migration is one every organisation faces at some stage.

Another key element of the process is communication. With a huge number of people impacted by the launch of a new ERP system, communicating clearly and frequently with them is critical to success.

University internal comms teams are often quite small and fully committed to delivering their business-as-usual commitments. This can marginalise comms and leave a great technical solution without an audience ready for the roll-out.

Planning:

Where to start is the easier of the two questions to answer - with strategy. Starting a project by taking the time to answer the critical questions for the migration is time and money well spent.

The strategy phase provides the direction for the rest of the process. It is where the blueprints for the whole project are developed, the scope and scale of the process are agreed and, importantly, the time impact on those involved is sketched out.

When to start is harder to answer, although ‘earlier than you’re thinking of’ might be the correct answer. Because the process of migrating data happens fairly late in any software purchase or merger of systems, the planning process often happens quite late on too.

But that’s one of the three main issues our experts spoke about that cause issues with poorly delivered data migration projects.

Conclusion

Data migration is a critical part of any transformation/data project, but is too often overlooked until it’s too late for proper planning.

But it doesn’t have to be this way. With a solid process, technical capabilities and resource planning, a successful, stress-free and accurate process is possible.

Universities are large and complex organisations that rely on data, and working with a partner who has the experience and expertise to work with you through this process can often be the quickest path to success.