“Data Literacy: It’s all about how to get people in an organization engaged so that they can make decisions and take actions that can collectively drive better decision-making,” said Harald Smith, Director of Product Management at Syncsort. Smith is a strong proponent of Data Democracy. “Data Democratization” is the idea that digital information should be accessible and understandable to the average end user as a basis for decision-making. “It involves everybody getting involved. That’s the overall push toward Data Literacy and Data Democratization,” he said.
Data Democratization has been promoted as a competitive advantage in the global economy and a desirable, egalitarian end-state where all decisions are data-driven. Smith talks about how organizations can encourage data literacy company-wide and how the position of Chief Data Officer (CDO) can facilitate the process of becoming more data-driven.
The CDO and the Proactive Mindset
In 2007 Smith attended a Data Quality conference where the need for a position beyond the traditional Chief Information Officer (CIO) role was emerging, and a popular topic was the concept of a C-level position that would intersect data with the business. About five years ago, he was hearing much on that topic where between 15 and 20 percent of organizations had begun to embrace the concept of a Chief Data Officer (CDO).
“At that time a lot of the attention seemed to be around Data Governance,” with the primary focus being regulatory compliance, much of it arising out of the financial crisis in the late 2000s. Smith said that selling the idea of a position centered on mitigation of risk and response to data breaches was not easy, until the idea of moving from a defensive, reactive mindset to a proactive one started to gain traction.
He saw a tipping point about three years ago when he speculated that “40 to 50 percent of organizations had added a Chief Data Officer.” Gartner predicts that number may grow to 90 percent in the next few years. He thinks 90 percent is a bit optimistic, but said, “We’ve hit that point where the Chief Data Officers are here and their focus is not just on Data Governance.”
Syncsort holds an annual customer advisory board conference and Smith sees a trend toward more CDOs in attendance as well as people from the business side who work closely with their CDOs. Their focus is expanding beyond a defensive strategy to monetizing data, innovative and disruptive activities, and facilitating data literacy. As an integral part of the business, he said, “That Chief Data Officer is really that center point for engaging across lines of business now.”
Role of the CDO
The evolution from a defensive or reactive culture to a proactive one requires raising awareness around data privacy and data security, and understanding the meaning behind regulations like GDPR. He said it’s about getting people to think about the data that they’re dealing with, and not treating it as something that’s just happening in the background over in IT, but rather, “This is something I’ve been dealing with day-to-day, year in, year out. This is something I’m responsible for.”
To that end, Smith has noticed an interesting trend: Chief Data Officers beginning to form their own product management teams. “They are asking, ‘How do we organize data so that it has a value within our organization, or potentially, as a commodity outside of our organization, based on what we’re doing in the business?’”
Developing a Data Literate Culture
Smith said the idea of data literacy began to emerge as companies started to see the value of Data Science and opportunities for innovation. The inclination at first was to hire more people from outside with high levels of Data Literacy, but when companies realized there were few Data Scientists that were familiar with the needs of their particular business, they decided to build teams internally, and Smith sees this as an important change in mindset.
People within the company who already work with data, who understand the workflows and how information moves, and who understand the domains of that data can be good first-level resources to develop, he said. Skills needed are literacy, basic math, ease with numbers, sequencing (whether of events, values, or even data processes), an ability to put data into a context, and a basic understanding of cause and effect. An employee working with spreadsheets and reports may ask what data was used for the report, how it was aggregated, and how a big number, for example, makes sense in context. “And then you’re at the point where they begin to actually do something with that information, and we can help people to understand that they have the skills.”
Next-level skills are a basic understanding of statistics, probability, and an ability to see correlations in the data, he said. In some cases, when someone has the skills, knows how to aggregate data, and has confidence in their basic skillset, helping them gain a deeper understanding of the domain they’re working in is the next step. “The core piece to help understanding is offering something like a business glossary that allows you to collect that terminology together,” he said. Learning the domain’s important terms, where they originated, and why they’re important can be facilitated using a wiki, a business glossary, or a SharePoint site.
Advantages of a Data Democracy
- Users in a Data Democracy are able to help with innovation: “A lot of good ideas are going to come from the ground up because people will say, ‘I could do this better’” said Smith. Users are suddenly able to see where to make quicker changes that have added value and have a sense of ownership for the solution that arises from their ideas.
- The CDO can support autonomy and self-service through education: “You’re in a position to help people explore data better,” Smith said, because users have permission to ask, “okay, how do I read this particular report?” or “How do I understand what this spreadsheet contains?”
- Users may discover new patterns that indicate errors or opportunities: A user may notice a column heading with an odd name and attempt to learn its meaning by looking in the Data Catalog. If the user discovers that the content doesn’t match the definition, they can bring it to the attention of relevant business teams, or the CDO’s office.
- Empowered users can spot areas in need of clarification: A question may arise such as, “there seems to be a particular outlier here. How do I deal with that?” providing an opportunity to expand a Data Catalog informed by how other people have used that field or that data. “That’s a second tier of communication that’s going beyond ‘What’s the content of that data and how do I get the understanding around it?’” he said.
- Potential security or trust issues can be found when more people interact with the data: A particular column of integration, for example, should have a particular frequency distribution log that reflects a statistical measure of how often it occurs. “I can look at that at the value level or at the pattern level, and as I begin to do that, then you can begin to figure out where you have potential issues of trust.”
- Decisions can be made on meaningful data across the organization: “That then ties into this broader story around getting everybody engaged with data, so that they can begin to take action and begin to make localized business decisions,” he said. Users learn to consult the data they have available to them, rather than guessing.
The Chief Data Officer: Facilitating Autonomy
Smith said that historically, access to data and skills needed to work with it was confined to a set of experts: programmers and Data Modelers and the Database Administrator, who served as gatekeepers. Business users had to ask for information, without having any understanding of what data could do for them. Smith remembers hearing, “I don’t really know what you have, but can you get me a report of it?” from users in his IT days. IT would try to understand what the users wanted, yet users didn’t know what was possible, or how to ask for it, and both sides knew that there was a lot of frustration with the process.
Now, rather than serving as a gatekeeper, the CDO can say:
“Okay, I’m part of the business side, but my role is to help enable you so that you can do it,” he said. “A Chief Data Officer can help engage across lines of business by facilitating that common language of the business, and that’s a very solid starting point to get people more data literate.”
Smith sees the position as a conduit for Data Democracy and a way to guide staff as individuals toward greater autonomy around data. Suggesting a data privacy course, for example, might engender a conversation about how privacy is connected to reporting processes, and the CDO can provide encouragement – and permission – to find answers independently. Being able to highlight a term on a report, look it up in the business glossary, and find the meaning of the term and the owner of the data empowers users, he said.
A CDO can establish effective two-way communication coming up from the bottom by encouraging users to ask questions, showing them where to start looking for answers, and facilitating that process. “This is how you begin to establish a data literate culture.”