Data strategy needs data governance

Ultimately, it all comes down to data. This is according to Chris Millet, a Baker Tilly director specializing in client engagement and managed services during a talk today at the Finance and Accounting Technology Expo in New York City. While there are powerful technology solutions available to professionals today, especially AI-based ones, none will perform their best without strong data sources paired with a robust governance framework. This is not only to avoid risk but to drive the organization’s strategic goals going forward. 

“We should be looking at our data as a strategic asset. It really allows us to transform our data into a competitive advantage for our organization, and really bring it into alignment with our overall business goals and strategy, and making sure that we’re looking at that data quality and governance, and really unlocking that potential in a robust way,” he said. 

He emphasized the importance of centralizing disparate data sources as policy, as he believes it is only when we are able to look at all the data together that we are able to understand the full picture of what is happening in an organization. When information is stored across separate silos that don’t connect or communicate, it becomes difficult to trust the system, as there is no single source of truth. 

“[If] none of my systems talk to each other, I can’t really understand what’s happening in this system versus what I am trying to accomplish overall and understanding my data. What is that single source of truth that I can really trust?” he said. 

Of course, even if one does manage to centralize one’s data sources, it does no good if the data itself is suspect. This is where governance comes in, and he said there are numerous solutions available to help. Regardless of which ones someone chooses, he said they should be able to automatically point out where data hygiene is less than perfect. 

“[It] can say, hey, you’ve got this many transactions that are missing this data point. You’ve got this many records that don’t have this piece of information that’s critical to your analysis. And so your tool should highlight those things and be able to direct you to where you need to fix your data, because it doesn’t matter what comes out on your dashboards or reports, or AI [if it is] garbage in, garbage out,” he said. 

While there are many ways to approach data governance, the very first step should be defining data ownership, according to Millet. Ask: who owns the data from what sources. Does the warehouse team own the operations data? Does the finance team own the finance data? Does the sales team own the customer relation data? Who actually owns the governance around the data that’s going to be consumed down the line? This also includes defining which users have access to which data sets, not just the inputs and outputs. This gives a true sense of both where the data is coming from and who is consuming it. 

He also talked about how, regardless of what solutions are deployed, they should also highlight compliance gaps. 

“And I’m not just talking about external compliance needs. You have third parties out there that may have certain requirements of you around your data, but I’m talking just as much or more about your internal compliance to go along with your internal data strategies and policies, and then lastly, enhancing the data integrity. When we understand there might be something wrong, [we can] go ahead and improve that and iterate over time so the more I can use my tools to help me identify the exceptions to my rules, the better and better I should get, and the more I can trust that data long term,” he said. 

Millet added that organizations should also make sure their data strategies and tools will scale as the organization grows. 

“We must be able to grow with this is not something we’re going to implement today, and then two years later we got to rip out and do something else. It should be something that I can grow with over time, and be able to be spread across multiple business systems,” he said. 

Finally, he also stressed the importance of stakeholder engagement. Implementing the most sophisticated, powerful solutions will mean nothing if employees won’t use it. 

“If [you’ve ever switched] to a new system, whether it’s NetSuite or another system, we’ve probably all heard ‘I don’t know what’s happening here, the reports aren’t right.’ Whatever the situation is, stakeholder engagement can only be there if they’re trusting what they’re seeing, and that the data strategy addresses that,” he said.