How do you go about maturing data governance and data quality in a large, global organisation?
In this interview we hear from Matt Smith, Data Quality and Data Governance Manager at Covidien who has been helping his organisation grow data quality and governance capabilities whilst delivering tangible results.
Dylan Jones: Matt, how did you get started in Data Governance and Data Quality?
Matt Smith: It was a case of being in the right place at the right time. I was working as a contractor in a variety of different areas from shared service start-up to project management to system implementations when a number of permanent roles were put in front me. The data quality manager for EMEA position immediately jumped out at me because no matter what area of the business or project I had been involved in I had always experienced the pain of poor data.
The opportunity to make an impact was clear, and it meant I had the opportunity to build something from the bottom up which was too much of an exciting opportunity to turn down.
Dylan Jones: How did you pitch your vision for Data Governance and Data Quality to executive management?
Matt Smith: I took a step back before moving forward by creating a bespoke data quality and governance approach for Covidien. It had to be a robust, holistic and adaptable process for how we would tackle the issue of master data quality and governance.
One of the most important decisions I made was to brand it; it became a product, a terminology, used internally when an issue arises. It is also independent of technology so its stands on its own but has the ability to integrate data tools to accelerate the process and improve delivery.
Everything we do is structured, documented, repeatable, measured and relevant. The resources we’ve created have already provided value to other areas of the business I didn’t anticipate during its design. We spend nearly as much time understanding our data as we do cleansing it and that by itself is a commodity which is commonly undervalued.
The word Data can turn people off before you’ve even hit your stride so I’ve always tried to relate what we’re trying to achieve to a different scenario (my favourite being the close relationship between data quality and governance with baseball). Making people think about data outside of the office environment somehow makes it easier to connect with the issue you are trying to overcome within it.
Dylan Jones: Can you briefly describe the journey you’ve been leading at Covidien?
Matt Smith: My initial role was based in the EMEA markets with a pretty general scope outlined as ‘Master Data’. By breaking this scope into manageable data domains specific to Covidien’s architecture and operations we were are able to target critical areas without ignoring the big picture.
By following our methodology we delivered tangible results along with a road map for continued success which we could clearly communicate to the business. This EMEA work led to increasing visibility throughout the global organisation and as larger projects were being deployed the need for a specific focus on data was identified and we were called in.
Having a strong methodology behind us meant we were able to hit the ground running and gain further adoption as various regions learnt of our approach. This organic adoption has meant the scope of the data attributes we’re responsible for and the geographical coverage has consistently accelerated.
Dylan Jones: You’ve achieved a lot in a short space of time, how did you learn the techniques necessary to lead a Global Data Quality initiative?
Matt Smith: A lot of research, reading and a bit more research…plus being prepared to take a few risks! I invested time at the outset to create something foundationally sound and appropriate for Covidien providing confidence that we could handle the natural growth of our responsibilities.
Too often the focus can be short term — ‘we need to clean the data for this shiny new application’ — but for me the value and excitement has always been in building a repeatable process. I wanted to create something that could be embedded as a way of working so it had to be well thought out and adaptable to requirements.
It’s also important to mention that I was lucky enough to be heavily supported and afforded the time to create and implement this approach. Again, having the support of a lucid methodology has been fundamental to what we’ve achieved, and it’s something I believe should be common practice for all data activities as it not only allows for a structured approach and clear deliverables but it also provides confidence to the business and sponsors.
We are constantly looking to improve and refine what we do as the techniques required have as much of a shelf life as the data itself.
Dylan Jones: What does the organisational structure for Data Governance and Data Quality look like in Covidien, how have you distributed the capabilities required to make this happen on the ground?
Matt Smith: Master Data was divided into seven key areas of focus: Customer, Supplier, Item, Price, Financial, Sales and Employee with roles created to be accountable for each domain.
By having ownership of a data domain we now have ‘Data SME’s’ as well as practitioners of the data quality and governance methodology. They’ve been able to take this skillset and apply it to the various challenges and opportunities that have come our way. By being responsible for a deep analysis of the data framework, cleansing activities and implementation of metrics, they have a truly holistic view of their specific areas.
This supports the governance model as their knowledge can be applied to a ‘Data Steward’ type role crucial to the ongoing governance and quality of the data. I’m fortunate that the individuals that make up this team are highly adaptable and skilled which is a major reason the work has been successful.
Dylan Jones: What is a Global Data Governance Model and how did you go about creating this resource?
Matt Smith: The biggest issue I found with Data Governance is that it means different things to different people – are we referring to governing data content or data rules…or both? The first objective has been to make the distinction and clearly communicate our focus.
The model is there to define the process, the various roles and their responsibilities in ensuring the data is appropriately managed. We have to make sure the right people are able to make the right decisions about our data. It has to be global in its structure but regional in its roll-out. Ours is fairly logical and can be applied either locally or globally (or more realistically I suppose, as a progression).
This is a journey and we still have a long way to go. But by defining our ultimate global model and then breaking it down to achievable and focused deployments, we’re able to build our approach out accurately and with a clear path towards our final goal.
Data governance is the glove to our data quality hand, and it is vital these two efforts are in sync. We try to see data governance as our end-game; if we’ve done everything as planned then our governance model is the obvious way of keeping our data in order.
About Matt Smith
Matt Smith describes himself as an entrepreneurial, strategic, data leader with an eye for detail and an appetite for structure. Experienced in multiple disciplines from top-forty musician to running a global .com (with over 40,000 members) before settling in the rock ‘n’ roll world of data.
His wish is to create and deliver successful, long term approaches to the undervalued world of data.
Find Matt on LinkedIn: http://uk.linkedin.com/pub/matt-smith/64/850/771/
Image Credits: Aaron Escobar, Flickr