The final piece of the puzzle
I kicked off this series with the statement ‘MDM – It’s not about the technology’. However, as I was writing, I knew that I would have to close off and talk about the technology, it is a key part of the MDM world, a critical enabler to MDM success. So, at the risk of sounding like a complete hypocrite and contradicting myself, let’s talk about the technology.
I must admit, I have been scratching my head on how I was going to bring this all together to provide a fitting end to this series. The content and ideas that I have written down could fill a book or two, so where do I start and how far do I go? How do I keep this blog entry interesting and on topic, whilst still following the theme of being ‘business focused’.
Luckily I have had a week of golf and sangria in the Spanish sun, which helped me get my thoughts together, let alone improving my golf game no end.
Technology is the tie in!
People, Process and Technology – The link
In the previous blogs I have discussed that to get MDM right you need to cover off the organizational and business side of the MDM strategy triangle before you focus on the Technology. And more specifically you should focus on the people and process areas of the business, with their relationship to each other and MDM.
There is method to my madness, focusing on the business and organizational aspects will provide you with a detailed steer for your MDM initiative. You will have an understanding of what is needed to align it with your business strategy and know what capabilities exist within your resource pool. You will now be able to define the MDM blueprint for your business, the definitive roadmap to MDM success.
Business Buy In
Another significant benefit is that having spent the time talking to the business, understanding the people and their processes, you will have already built a bridge to a major factor in MDM success, business buy in. The organization will feel part of the team, and they will believe that you care about what their issues are and that you want to help them.
Business change management is a critical factor in any MDM deployment, getting their buy-in is essential.
However, I digress, let’s get back to why Technology is the link!
Having defined your strategy and roadmap for MDM, you will have identified a capability gap between what is required and what is currently provided. This gap will consist of:
1) Missing / undefined and unmanaged business processes
2) Lack of Data Governance processes and Data Quality management
3) Skills gap on the front line and in IT
4) Technical capability within IT in the provision of MDM
5) Gap between current state data usage and provision and your MDM enterprise vision
These are examples, there will be many more items in this list. The key point here is that without the right technology you will not be able to close the gap. AND, having developed a sound business focused MDM strategy, you will be in the position to select the right technology, a technical solution that is fit for purpose, in your current state and scalable for the future.
You are now ready for Technology – BRING IT ON!
What do I do now?
When I embarked on my current MDM deployment I thought, “This will be easy, it’s just like another data warehouse deployment, I’ve done tonnes of those!”
WRONG!
Well, not quite, there are a number of similarities in approach, from data discovery to requirements prioritization and iterative deployments. However to deliver true enterprise wide MDM that is part of your business and technology DNA, you need to take a wider view. I liken it more to an enterprise application selection and delivery, it brought back memories of when I was an ERP consultant, and some of those memories I had suppressed on purpose! Oh the nightmares …
For MDM you need to take an enterprise approach to deployment.
Conclusion
I am not going to go into how to mange an MDM program / project in this post, it’s a series all on its own. However to get things started I will finish with the following advice:
- Ensure that you have executive buy-in to your MDM blueprint
- Select the technology that fits what you need, don’t go too small or too big
- You need to be able to work with the vendor or the consulting house, MDM is not an out of the box solution
- Run a detailed proof of concept, with clearly defined success criteria
- Take small steps, don’t go big bang
- Don’t throw everything into your Hub, or risk it becoming a dustbin for your data
- Ensure it is scalable, if MDM is a success in your business, your business will grow, and so will your deployment
- Don’t be afraid to take risks, stay ahead of the game
- HAVE FUN!
I like the last point, you need to be passionate about MDM, get it into your DNA and you will go a long way towards making it part of the DNA of your organization.
I’ll say it again, HAVE FUN!
PS: To help with the next steps have a look at http://www.dataqualitypro.com/ and feel free to ask questions of their community of experts.
Excellent series Charles,
I really enjoy your straightforward writing style.
Covering the People and Process sides of the MDM strategy before selecting the Technology is excellent advice.
As is having fun and making MDM part of the very DNA of the organization – I am a big fan of TLMs (three letter acronyms).
I am also a fan of top ten lists, so perhaps we could add #10 – Go to Spain for some golf and sangria before launching your MDM program?
Best Regards…
Jim
Thanks for the kind words Jim.
I agree, we should kick off all MDM initiatives with a week of golf and sangria in the sun, can only help …
Cheers
Charles
Charles, your ramblings in MDM have been very thoughtful and actually very structured.
When we talk about the holy triangle of people, process and technology I am always a bit skeptical when we start selecting one to be more important than the other and making none iterative sequence with these terms.
In my eyes aligning people, process and technology is the Holy Grail we should be searching for.
This means you may use technology in early phases. Data profiling technology is often used as an exercise in early phases in order to understand your data and bring forward hidden business rules or light up business rules which are wrongly thought to be in place.
Such data profiling may include use of external reference data also.
I am actually right now doing some data discovery with client data. A population of customer records marked as ‘business’ entities are matched with the national Danish company register. Surprisingly only slightly over the half of them seems to exist in the register given a confident similarity threshold. Mama Mia – do I have bad algorithms?
Well, it was actually the population without an attached national ID I was working with, so it is always difficult working with the known dirty segments.
But – learned from past experiences – I also make a match with the name field and the list of person names in use in Denmark from the statistical bureau. Bingo. One third of all names match that list and gazing the addresses I have a feeling they are private residence addresses.
I do have some more technology and reference data to go with like profiling the addresses with the national building register (cost applies) and make a specific match with owner data in the national company register. But for now I will discuss the results with the client – probably it is some kind of business rule that makes the ‘business’ assignment to customers look different than expected.
And huh – I see a lot of duplicates in there.
Thanks very much for your comments Henrik. And for sharing your experiences, much appreciated.
I have to agree 100% that it is the alignment of people, process and technology that is the holy grail.
I suppose that what I have been saying in this series is that if you understand your people and process first the alignment will be a lot easier.
As you say though, the use of technology to get you to a point where you understand and know what the gaps are, especially in 'process', can not be ignored. Once again, the technology is a critical enabler in this and provides the essential capability to get you to that understanding.
Believe me, I can't live without the technology that I deploy, it is key to my success, and the success of any MDM / Data quality / Data Governance initiative.
Hi Charles,
Good series! I may agree with you as an ideal scenario, but I don't think that is feasible most of the time.
If you're building your MDM solution, maybe your approach is best. However, if you're buying, I don't think so.
Regarding MDM solutions, the pendulum is shifting from building to buying. As such, it doesn't help if you figure out everything about people and process, and then you can't find the right technology to match it.
I say you need to have some basic requirements first to drive your technology selection, and then customize your particular needs on top of it by adjusting people/process/technology along with it.
Hi Dalton
Thanks for your comments, and the debate.
I hope I am reading you right, are you saying that you should customize / adjust process to fit the technology solution?
If so, is there not a massive change management risk there, i.e. you run the risk of changing core business process because of gaps between the technology and your business.
Would it not be better to understand what you need from a business (process and people) point of view and then get a technology that is customizable to fit your requirements.
Re the customizable point here, I don't believe that you can get any MD technology / application that can be deployed out of the box, even in a greenfield environment, therefore you will have to undertake an amount of customization at the least, thus you have the opportunity to get the technology to fit your business.
You're reading me right, Charles, and you're right: a comprehensive change management process is a must.
Take Sun, for example. It would be practically impossible to reach large scale consensus on what the business "need." Each line-of-business would have a different view. You get to a point that you have to draw a line, and that's why Data Governance is so important too.
Yes, I agree just about every solution will need customization. And that's my point. You need to find, based on some fundamental requirements, what solution will be able to "grow" with your needs.
In the end, it is a compromise. Technology will have to adapt to process, and process will have to adapt to technology. And that is, in my mind, an evolving process. You can't expect to have one side of the equation totally finished before you decide on the other.
Dalton, with you 100% in that you won't get large scale consensus at a detail level across the business, and you shouldn’t even try. However, taking the approach of gaining high level understanding of the business needs and building the MDM blueprint from those needs, provides a roadmap, defining clear subject areas and their appropriate benefit and priority, that can be sold at exec level, an ideal tool for getting business buy-in. This is an approach that I have used in MDM and data warehousing programs as well as large ERP change programs. What you then do is plan iterative developments off this roadmap, getting into the detailed prioritisation and planning at a subject level, as well as map out what your data governance processes and structure should be.
I have been blunt and singularly focused in each post in this series, to ensure that each point of discussion held centre stage. But with your points, as with the points made by Henrik, aligning people, process and technology is key to success. I will stand by my belief that you need to understand your business process and people first, technology should be there to help you achieve those goals. It’s the selection of the right technology that is key, and you will need to know your business needs and capability before your know what the right technology is. Maybe I just don’t like the idea of compromising
Isn’t it great that we have a community like this, of successful and highly experienced people from around the world, and we can still debate approaches like this. What a blast!