Caring is crucial in order to manage relationships as well as assets – we care about the things that matter to us. You know this from your private life. If you didn’t care about your most valuable assets, such as your kids, spouse, friends, car, boat or whatever they may be, it’s not hard to imagine what could happen: your kids would go totally out of control, your spouse and friends would leave you, your car would break down and your boat would sink. The same goes for any company – private or public – across any industry: we need to manage the things we care about, or to put it differently, we manage our assets.
So what are the important assets in your company? Most likely they are employees, products, buildings, capital, infrastructure, etc. These are all fairly tangible assets in the sense that we can hire and retire employees, we can buy and sell products, and we may invest in various markets and products and so on. We see dedicated ownership of all of these assets through various senior executives in finance, HR, IT, manufacturing and other business units. But who owns the data? If you, as many others do, consider data to be the product of your processes, data should be one of your most important assets, if not the most important. Additionally, data is one of our longest sustaining assets, since it was there yesterday and it will be there tomorrow. You don’t dispose of data that represents value to your business. People in the organization say they care about the data, but do they really? If caring means to focus, to be responsible and accountable, to define the one version of the data that meets the enterprise’s needs, then odds are nobody in your organization cares enough about data.
If the need to care for data and manage it as an asset is so obvious, then why isn’t it happening? Why isn’t anyone volunteering during the annual shuffle of job descriptions and responsibilities for the year to come?
There are a few possible reasons. First, nobody is asking anyone to take care of it because nobody realizes this should be treated as its own task to add business value. Secondly, nobody voluntarily takes on the task of ensuring data quality and consistency because they believe it’s a “mission impossible” that is bound to fail, and they don’t want to be part of that failure. Or thirdly, it could be that many people within the organization know something must be done, they just don’t know what it is, how it should be done or who should take care of it. These people (the stakeholders) are typically found within business units feeling the daily pains related to their specific unit and have related these pains to the issue of data.
Since the pains are fragmented, and since no one “owns” the cross-business unit processes, then the pains rarely get big enough to reach top management’s attention, therefore, no initiatives are launched to fix them, which results in resources being thrown at the pains to mitigate them locally. The stakeholders could also be members of enterprise architecture functions that orchestrate commonalities and create synergies through standardization and professionalization, but they don’t get the attention required to make it happen because they are firing on too many cylinders.
I have met many organizations claiming they care about their IT, thus their data. Some of them have even defined dedicated roles for caring about the data, e.g., data stewards or custodians who are typically owners of data within a system. However, that doesn’t do the job. It’s a start, and it’s a part of the solution, but only a small part of it. Data in itself is not an IT issue. Storing the data is an IT issue, but understanding and interpreting the data is a business issue done by the business functions. This is one of the primary challenges when trying to place the ownership. Data is not owned, but partly managed by IT, and business doesn’t want to own anything that “smells” like IT. This is typically the reason ownership is not anchored, and hence, falls between two chairs.
So what does it mean to care about data? I call the process of doing so data governance. This is all about placing ownership, responsibility and accountability of data representing critical business entities that really mean something to the business, i.e., we are not talking about all data across the enterprise, but merely certain named entities typically used across business units, across borders in an international organization or across critical business processes. These so-called entities could be products, customers, vendors, employees or other relevant business-critical objects and are also seen as master data entities. The purpose of the limitation is to improve prioritization of the work at hand to create business results, since the task of trying to govern all data in the enterprise is simply not feasible, and will most likely not be beneficial either.
The work involved in defining and implementing data governance in an organization is not trivial, and it requires time and money to be successful. Also, it requires organization-wide acceptance to adopt and implement the changes into existing or new business processes. Based on my experience helping organizations benefit from data governance, the work involved activities such as:
- Creating the strategic link: Ask yourself why you need to do anything and what the expected value is. Identify the link between doing something and your corporate strategic objectives as defined within the company’s strategy.
- Deciding on scope and ambition level: Decide what is in scope to be governed. Is it local data and enterprise data? What role is data warehouse data playing? Which systems are involved?
- Identifying entities to be governed: Which data entities are in scope, i.e. customer, product, vendor, employee, etc.
- Modeling the entities: If you can’t model it, you can’t manage it. A prerequisite for defining the appropriate governance model is to design entities including the basic business definition, core minimum attributes, hierarchies, the life cycle and relevant classification.
- Identifying stakeholders: Who should be actively involved and why? Which roles are they likely to take, and how does it match the current organization in terms of their current positions and related power?
- Defining the data governance organizational model: Draw the governance model including all identified levels, roles and stakeholders from an enterprise level down to the local business units.
- Identifying and implementing changes in business processes: Document the new governance processes and how they either supplement or replace existing business processes.
- Communication, communication and communication: Constantly communicate the purpose and value of governance in order to change the culture toward adopting the new data governance organization and processes as a natural part of their every day duties and ways of working.
It is my sincere belief that data governance is the single most valuable activity in any master data management (MDM) initiative. Even without deploying any supporting MDM technology, data governance will add business value if deployed correctly within an organization.
Therefore I recommend that you care about your data as an asset to your organization if you want the benefits of having it. There is no need to start from scratch with a blank sheet of paper, since there is plenty of experience out there, as well as people and companies who have invested in documenting the best practices for organizations globally and across industries.
Martin ABC Hansen serves as executive vice president at Platon A/S, where he is responsible for the company’s international business.