Sometimes it is easier to go with the low hanging fruit of financial data / meta data, but as time grows, the benefits outside of finance (like your Maintenance example having to know more elements than finance) can be clearly articulated. Imagine the start of an MDM project with one focus, then growing that data model to include other areas of the business like Maintenance or shipping, channels, vendors, retail locations…

Thanks for the clear example, scenario. Really drives home the value of MDM in our world.

It’s an excellent question, Jackie. Maybe one day I’ll be smart enough to actually to know the answer.

Give me an older system with good data over a newer one with garbage any day of the week and twice on Sunday.

Expectation management is one of the challenges of “new systems” and new “software products”. The marketing material, and the business case always highlights what the new system is, undoubtedly, CAPABLE of doing.

The marketing demo shown to the senior execs, always uses pre-populated sample data, to illustrate the wonderful capabilities of the new system.

The fact that a new system is CAPABLE of accepting perfect data, and CAPABLE of generating wonderful results, on the basis of perfect data, is no indication of what the new system will ACTUALLY do.

Most new systems need to be populated by a Data Migration project. Failing to plan data cleansing of the old data is shockingly common place.

Then senior execs express surprise when the “new system” fails to live up to their expectations….

Some “new systems” are required for regulatory compliance reasons (Anti Money Laundering, Solvency II, etc. ). Regulators are beginning to seek evidence of the data population processes – I cite some examples on my blog.

Rgds Ken

I think people take data for granted like we do hardware. Not to mention, how often and long do people talk about the UI features they want? It’s more tangible to business users, somehow?

In the end, it’s always going to be there and we need to be slick in the ways we try to prevent it and clean it up when it does happen.

I completely agree with your reply to William.

The attitude too often taken is: “Our job is to implement the new system, not clean up someone else’s mess”.

Ken

2. should verification at source (contact for address or manufacturer / supplier for product) be required at initial setup of the data in the system or maintenance scheduled as part of the data governance program? – yes, otherwise what are we governing and by what laws/rules.

3. Is the data incomplete? – yes, often address data is incomplete or invalid. Either is insufficient to become information.

4. Does the MDM process include a question / answer scenario to complete the data? – I believe so. This is the voting portion of the governance initiative, functioning as the business’ chance to manage its own information.

Great post Jackie!! Keep it up! I’ll be mulling these questions over and forming a blog post soon!

The classification of maintenance items is a challenge especially when you are working at multiple manufacturing facilities where the terminology and classifications are localized. We set up the standards, provided a clean classified and attributed record verified to the manufacturer of the component to the plant which improved our ability to standardize the MRO data up front. Part of our work with the maintenance and engineering departments was to set up synonym search capabilities which enhance the ability to find a part.
Jackie

Jackie

Jackie

DaveP

Before considering any form of “Transformation”, the “business rules” for the field should be fully understood:

I would want to have:

1. Business name of the data field(s):
2. Permitted values:
3. Business meaning of each permitted value:
4. Interdependencies with other data:
5. Field precedence:

The “business rules” should be centrally maintained, and visible to the business.
All “Transformation” rules should also be centrally maintained (and not just coded into an ETL script).
Actual transformations, and exception processing (when input data could not be transformed) should be recorded in an audit trail.

Regulators will increasingly look for a “Data Audit Trail” – be warned.

Ken

Count me in too.

I will be happy to share real world examples from my experience in requirements definition, sourcing, migrating, profiling, structuring, mismatching and re-verifying legacy system data.

I look forward to learning from the experience of others. Working together we can achieve far more.

Ken

I hear you.

One of my goals for Data Quality Pro was to create a channel for the many practitioners who can’t find the time to get to conferences or even blog about how they are delivering these initiatives on the ground.

I think the DQ/MDM/DG blogosphere has really expanded in the last 12 months but I agree we could do with a lot more case studies and practical accounts.

Shameless plug coming up…

I’m launching http://www.DQDirections.com late in Q2 this year to provide a virtual presentation platform for many of these case study organisations, many of them I’m sure will find their way out onto the blogosphere in the form of interviews and podcasts.

In the meantime, I’m also keen to gather and collaborate on more “how” stories, theory is great but I agree we need more stories from the coalface.

To pick up on one of your questions – classification of maintenance items, I have a real world example that demonstrates just how challenging this can be!
On a recent project I needed to develop the requirements for an asset register for a major rail project. The complication was the fact that different organisations would be managing the assets on completion of the project. The two main organisations have different asset classification structures which are totally incompatible. Due to their different origins, they reflect different classes of assets in different ways, have different levels of granularity, and different ‘branches’ in the tree structure. Additionally, both organisations are continuing to refine their classifications on an ongoing basis, with no synchronisation and no driver to adopt a more consistent approach.

A further complication is that the CAD team use the Uniclass coding system, which is fine for buildings etc. but is light on the particular assets required.

I was able to develop an approach which all parties are happy with, but this did make the overall process more complex than a typical infrastructure project.

Hope this example is of some use.

Julian

I’m in! As the name of my blog implies, The Data Quality Chronicle was setup to do just as you propose; get into the details of how data quality is implemented. With regard to the specific questions you asked above, I’m most experienced with cleansing/profiling and implementation of CDI with CRM packages (Microsoft Dynamics CRM to be specific).

I look forward to the exchange!

William

No doubt that we need discussions on why data governance, master data management and data quality is important – and sure we actually also have that in the blogosphere.

Also we need to exchange ideas on how to establish ROI for these disciplines. That exist too.

But we certainly also need to share concrete opinions, experiences and observations on exactly how to achieve improvements – may they be of general nature or be specific to the single entity types and lines of business.

This will actually help in going round the circle. If we know how to solve the issues we will be much better in clarifying the investment part of the ROI. In doing that we will see that most people actually already know that data quality etc. is important, but many people, not at least senior management, don’t want to go to action if there is no facts for how to succeed.

In the realm of data quality tools and services there seem to be some trend of having it look like a kind of magic sorcery. It is a kind of an art – but as you have pointed out, it is mostly a question of ensuring efficient knowledge driven business processes, having the right data classifications and metadata definitions, activating translations and so on.

There is a whole world of real world examples we can share.