5 lessons I’ve learned the hard way when implementing master data systems

I think I’m not exaggerating when saying that Master Data Systems (MDS) have a boring image. No fancy user interface, no AI, no nothing. Not sexy at all.

And I’m not exaggerating when stating that many MDS projects lasts much longer than expected, don’t achieve set goals or fail completely.

Personally, I think this is a tragedy. Although MDS are not sexy, they do provide tremendous benefits to airlines in specific and enterprises in general. Enhanced process efficiency, reduction of workload, improved process quality, etc. However, it is not the purpose of this blogpost to convince you about the benefits of an MDS.

What I’d like to write about are my personal lessons learned when implementing MDS. To be frank, I’ve encountered almost all of the above issues:

• Projects took much longer than expected
• We finally implemented an MDS but nobody wanted to use it
• The project did not meet the set goals and had to be re-engineered
• And I can tell you, it’s horrible. You are demotivated, the client is unsatisfied. It’s a huge mess — of course not always, but sometimes. Personally, it drives me crazy if a project does not achieve 100%.

We sit together after projects — internally, with the client and made up our mind, what we can improve. And when analysing each of the projects retrospectively, we could identify 5 crucial aspects that strongly influence the project’s success. 

Interesting enough, none of them is related to technology or the system itself. But you will read more about this later.

Actually, I’d call those 5 aspects essential for a successful master data project. If you are planning to implement a MDS you should invest five minutes to go through. It will save you a lot of time, money and stress in the end.

So, here we go with 5 things I’ve learned the hard way when implementing an MDS:

It’s a business project – not an IT

Let us directly start with the most important point: Implementing an MDS is a business topic. It has to be driven by your business department. They have to understand the concept. They have to understand the benefits. They want to have an MDS and on a higher level a master data management. Yes, it’s called Master Data SYSTEM and systems implies IT — I know. However, an MDS is just a small part of Master Data Management. Master Data Management contains so much more than just the system. 

So, without your business departments and a project champion from business (side) your MDS project will definitely fail. And fail in this context means, you will maybe implement an MDS but most likely no-one will/wants to use it.

Lessons learned #1: Get business departments on board and never ever start an MDS project which is driven by IT.

Don’t underestimate the change your initiating

As mentioned, an MDS is just a part of an overall Master Data Management. Implementing Master Data Management means a huge change to your organisation. Processes have to change. Responsibilities have to change. Workload will be shifted. Workload will vanish. Information silos have to be bursted.

I’m not telling you a secret that employees will be reluctant. Nobody wants change. Nobody want to lose responsibilities and nobody wants additional workload.

An MDS is not only a new system, it stands for a complete change in processes in an organisation — that is often forgotten or pushed aside. 

Lesson learned #2: The whole project has to be accompanied by a proper and well-developed change management process. If you can’t do that on your own, go for external change consultants. Second aspect in this context: Get a management buy-in. That’ll help you to enforce the changes.

Establish processes

Already said that earlier: MDS is about processes. Who is responsible for data? Who is allowed to make changes? When does responsibility of data shift from department A to department B. Just think about an aircraft as master data entity: There are dozens of departments which contribute master data during an aircraft lifecycle. Initially starting with strategic planning, network planning, maintenance planning, tactical departments and finally operations control, etc. And all of them have specific master data, absolutely vital for their processes.

Lesson learned #3: Master data processes are eminent. I’d say you should spend at least 50% of the project with defining processes and setting up guidelines. This is a lot, I know. But what’s the alternative: A system nobody uses because it’s unclear who has to add data, who is allowed to make changes and who is responsible.

Establish data governors

Operating an MDS and maintaining a Master Data Management is definitely not a one-off. It is a continuous process which has to be maintained, that has to be optimised and that has to be monitored. Otherwise you’ll end up in a complete mess.

Lesson learned #4: Establish data governor positions. And I’m not talking about a data police, searching for mistakes and prosecuting the authors. I’m talking about dedicated staff which are responsible for the further development of your MDS. They give advice, they are responsible for changing processes, they are responsible for adding new data and connecting new systems.

Content is king

Finally, I’m going to tell you a secret: The best MDS won’t be used without content. Ok, that’s obvious — but so important. Content is king. Without high-quality content and a broad range of content your MDS is useless.

Lesson learned #5: It’s never too early to start with creating content for your MDS. And always remember there are two sources for your content: Internally and externally. Just think about airport information (IATA codes, longitude, etc) or time zones, airline information, currency, etc. — they are all available and will be huge asset of your MDS.

Conclusion

Once we identified those crucial factors and have been able to internalise all of them, we’ve massively boosted our result quality. Certainly, we still encounter obstacles and hurdles in course of a project. However, having understood the core aspects, the essentials that are influencing the project success, we have been able to deliver MDS projects much more efficient and at higher quality.