What is Master Data Management?
![](https://idm.net.au/sites/idm.net.au/files/guyholmes.jpg)
I am hearing much about Master Data Management (MDM) from all industry sectors. Not just oil and gas but commercial organisations, software providers, miners etc. On my quest to be able to talk intelligently about it over dinner with my nerdy database friends, I decided to create “MDM for Dummies”. What follows is not for dummies, but for you – the intelligent and worldly reader.
Master Data Management is the policies, standards and tools for consistent management of an organisation’s data. The basic goal of MDM is to provide one version of the key business records – a single point of reference. It includes the policies and processes on managing and accessing the data from or within that single source.
PPDM vs MDM
MDM is a concept or “way of life”, whereas the PPDM data model (created by The Professional Petroleum Data Management (PPDM) Association) is far more tangible. PPDM provides the place to live the MDM dream. PPDM frequently exists where MDM does not. Conversely, MDM can never exist without a database built on PPDM or a similar specification.
MDM seeks to manage all key business data across the enterprise in a single point of reference whereas PPDM has restricted its model to areas relevant to the E&P industry.
What industries does MDM apply to?
The MDM concept is relevant to any company, whether they sell guitars or explore for oil, whereas PPDM is limited to one industry. Because MDM is a way of life, it can be applied to any business.
Consider a company that sells guitars. Their core business information might consist of the models they sell, their customers and addresses. Without a consistent system and process of managing this fundamental information, they won’t be selling many guitars or be in business for very long. It is easy to see the wide reach that MDM might have.
A PPDM database performs the physical function of storing core business information. Let’s consider well data. PPDM enforces a unique well identifier, a set of coordinates, total depth, etc. for each well. These foundation blocks are critical to a myriad of data and metadata that is related to this well. If information is wrong or incorrectly duplicated elsewhere, the flow-on effects can be massive across the organisation.
What data is considered “Master Data”?
Much of the data we have in our businesses probably does not qualify as master data. Well production statistics and the location of a drilling rig in transit are more transactional in nature than master. The key difference is that master data elements tend to repeat in other data sets, and therefore consistency is key to quality. Getting these building blocks right and tightly controlling their policies provide the backbone for many other data sets.
The use of the PPDM model itself has had the same obstacles as one would find getting an MDM system in place and management buy-in for changes to fundamental systems are never easy to get. I believe this is less from the cost of doing it, and more from the distraction and disruption that management sees in implementation.
PPDM has managed to convince software vendors in the oil sector to use the PPDM model as part of their underlying software solutions, and this has meant that companies end up using the PPDM model because they purchased a software product that uses the model under the hood (a nice bit of stealthy work from PPDM).
However, I do wonder how many companies then end up using the underlying database structure that PPDM provides as a sort of MDM central repository? How often do people using OpenWorks link the database to the Petrosys PPDM Compliant Database so the two applications share the data? In my view, the decision to join and maintain and manage that share is the start of the MDM style of thinking. In fact the decision to integrate these two databases is fundamental to the whole MDM concept.
How do you get MDM buy-in?
It is probably safe to say that most medium to large oil and gas explorers would benefit from MDM. However, the problem most of us face is convincing management of the need and value of an MDM solution and more importantly the MDM way of life.
MDM is a great idea and using PPDM as part of that system is even better. But as we all know, getting approval for something as game-changing as an MDM policy and tool set is where the whole thing starts to get shaky. If you want this to succeed, biting off more than you can chew is not a good idea. This is definitely an “eat the elephant one bite at a time” moment.
As with any project, start with a whiteboard, a few work colleagues who share the passion and a strong cup of coffee. See what a small scale project might look like. Find something with a significant impact that management will see. If they don’t understand the value proposition, traction will be slow. It must make their life easier or solve a recognized business problem.
Ideas for a good MDM starting point
Let’s say an oil company intends to explore in North Africa. Start the MDM process rolling while the data is being assembled. It is a time of few preconceived ideas about the data and how to assemble it; everyone wants to just work with the best available data. MDM with a PPDM backbone would help resolve the usual disparate data sets that start to evolve as drilling, regulator compliance and production each create their own operational databases. Help them all, show impact and make your life easier (in the long run) by getting them to use an MDM source for the key data sets in their projects. It is far harder to rope them back into a system you make if they have already started their own.
How do you ensure an MDM project succeeds and more importantly gets noticed?
Ensuring an MDM project’s success is no different than any other project. Organise the project including scope and stakeholders, plan the work and associated schedules, set achievable objectives, gain management buy-in and manage for success.
The one thing about MDM that can ruin your chance of recognisable success is that it can solve problems by stealth. No siren goes off when you get to the end, and the organisation will likely not see an immediate game-changing impact. MDM is more likely to make long-lasting differences that benefit many, get noticed by few, and reward management and the business more than the individual who kicked off the initiative.
Guy Holmes is a Sr. Vice President - Asia Pacific - Katalyst Data Management, Founder SpectrumData, DataReSource, and BeyondCool, with extensive experience in the oil and gas, minerals, medical, and information management industries in Australia and the Asia Pacific.
This article has been reproduced from the recently published Volume 2, Issue 2 of the PPDM Foundations Journal for 2Q2015 which can be viewed at the link here: http://ppdm.org/ppdm/PPDM/publication/PPDM/Publications.aspx