Breaking down a Data certificate…
This concept is debated around corporate board rooms, around conference tables surrounded by operational professionals and around the proverbial water cooler. But “data” doesn’t come with a certificate of ownership or a title like a car, boat or any other tangible asset that you own.
So then, what is data ownership?
Data ownership, according to Technopedia, is “primarily a data governance process that details an organization’s legal ownership of enterprise-wide data.”
This usually means that one organization within a large enterprise can create, edit, modify, share and restrict access to data or a dataset.
Often this will also mean that the “owner” can assign, share or surrender the rights of their data to someone else, a third party. This one fact has many large corporations wary as the General Data Protection Regulation (GDPR) in Europe starts to gain momentum.
Having anyone share information with a third party can open corporations to penalties, depending on the type of information shared.
With that risk in mind, data ownership inevitably leads companies to consider “accountability” for the data.
If you own the data, it logically follows that you are accountable for it. But in large companies the concept of ownership often stops with the application owned and managed by a department or function.
The data flows from one application to the next and the “ownership” goes unnoticed or managed UNTIL something happens to disrupt the flow of data or to cause the data quality to deteriorate.
This fluidity of data from function to function leads us to consider who actually is responsible for any single data element. Let’s imagine that we are the Marketing department of a large F1000 company.
We are responsible for data that flows through our marketing strategy and planning systems and, at some point in that process, data must connect into the sales function and the systems used by salespeople.
But imagine that marketing has taken “ownership” of the data that describes “customer.” This means that marketing is responsible for the “customers” to whom they market. And marketing’s primary concern is for the name, job title and email address (along with proper permission) of the customer who works for a company that your strategists target.
When that description of a “customer” flows to the sales team, will the salespeople feel that those data elements secured and owned by marketing are enough to help them sell a product or service?
So, one data owner (marketing) passes off to another data owner (sales) a customer record that may or not be helpful.
Sales data will, no doubt, need to be embellished by connecting contact information to a company record that details, for the sake of invoicing, the company address as well as other information for credit and taxing purposes.
Perhaps that you can now see what “ownership” means. The functions “own” the data if it is being used inside their portion of the business process.
When that data moves to the next process, its ownership also changes and herein lies the rub. Who then is “accountable” for the quality of the data as it moves through the data supply chain?
Usually the answer to that question is everyone and no one.
Everyone should be accountable for data that they touch in the normal course of a company’s daily operations, but the sad fact is that no one really feels responsible for doing so because they usually do not make the connection between their lack of perfect data entry and the downstream effect on others such as on billing, sales or customer service.
Like the people in Flint Michigan whose water problem finally surfaced as a major catastrophe in 2015, a company’s data issues are often below the line in budget and concern until the issues created are catastrophic.
The “owner” of the data in question must reckon with the accountability for imperfect data. The owner usually argues that the data was created appropriately within the process where it originated but was transformed downstream by other users.
Here is where ownership and accountability come together.
True ownership of a data element means owning it throughout the process whether it is transformed or not. Accountability means that the owner will stand up, acknowledge their responsibility and fix the problems.
Since ownership and accountability lie at the heart of data quality.
Here are three things to check on within your organization to ensure that you have the right emphasis on ownership and accountability:
1. Identify Data Owners.
Do you have a list of all the data elements that are most important to your company?
Start with customer, but include elements like product, partner, install base, financial hierarchy, sales hierarchy, employee. Assign an “owner” to each of these elements.
2. Identify Roles and Responsibilities.
With the list of data elements in hand, who is the person responsible for the well being of that data element?
Responsibility for that data element should include the budget to get and keep the data clean, the process that creates the data element and the metrics that measure the quality level of the element.
In addition, the responsibility for this data element should also include collaborative handoffs of data from one process to another.
In other words, if you are responsible for the creation of “customer,” you should also be responsible for how “customer” information flows from your process into the next – for example from sales that creates a “customer” record to finance that processes an order.
3. Manage data with data.
This is the point that counts. With owners in place and the data flow through a functional process well documented and agreed to, the next step is to ensure that you keep the data flowing efficiently.
That means that you need metrics.
Just as the foreman on a manufacturing line monitors the process for developing a car, someone must monitor the flow of data from one point in the process to the next. Most data errors are caused or identified at the points within a process where one application or one process hands off to another.
Just as in a relay race, the passing of the baton is the most important single task to get right if you hope to win the race. That’s why metrics for duplication, completeness and consistency are so important at these junctures.
Although your data assets may not have an official certificate of ownership, you can ensure that you have the processes and the people in place to make ownership real for those who depend on the data quality in your company’s systems. And with ownership comes true accountability.