What role have Data Management and Data Architecture played in data-driven organizations, especially during the tumultuous and uncertain period at the beginning of the COVID-19 pandemic? Industry thought leader Donna Burbank, the Managing Director of Global Data Strategy, discussed these issues in her presentation Trends in Data Architecture at DATAVERSITY®’s Data Architecture Online conference last July.
Burbank shared insights from a 2020 survey of almost 300 industry professionals from over 35 countries. A number of trends emerged, including insights into how organizations were seeking a safe “back to basics” that returned to the foundational elements of Data Management and the role of business intelligence and analytics. But first, Burbank needed to explain the difference between Data Management and Data Architecture, and how those concepts support each other.
The Three Pillars: Data Management, Data Architecture, and Data Governance
DAMA International’s Data Management Body of Knowledge (DMBoK) provides definitions of Data Management and Data Architecture:
- Data Management is “the development, execution, and supervision of plans, policies, programs, and practices that deliver, control, protect, and enhance the value of data and information assets throughout their lifecycles.”
- Data Architecture is “fundamental to Data Management … Data Architecture artifacts include specifications used to describe existing state, define data requirements, guide data integration, and control data assets as put forth in data strategy.”
Clearly, the two concepts are closely linked, as one definition includes the other. As Burbank explained, Data Architecture, Data Management, and Data Governance are “pillars that work together” to form a successful Data Strategy.
“Data Architecture is really truly more of the framework of the house,” said Burbank, giving examples such as understanding how on-premise databases and cloud fit together and having structure like a data model, data lineage, and a data flow diagram.
Building that solid foundation for a house takes time and effort upfront, but doing it right ultimately saves a lot of time in the end. “More people are looking at, ‘Is the data right? Is the data trusted? Do we have the right seat at the table for even what these metrics are at? We agree with how you calculate total sales, etc.?’” said Burbank. “It’s hard to do any of those above things without a strong Data Architecture.”
Data Management is a broader concept that looks at the alignment between business goals and technology, and then Data Governance is a still broader concept, which really deals with the people who are going to be working with the data: the culture of an organization and the processes, procedures, and policies around data.
“They’re all related,” Burbank explained, comparing the relationship of these three concepts to “a triangle with Data Governance at the top, then Data Management, and Data Architecture.”
Business Insights and Analytics: Key Drivers in Data Management and Data Architecture
While the pandemic certainly turned the world upside-down in many ways, there were still many consistencies with how organizations had responded to the survey in previous years.
When asked, “Which of the following have you already implemented in your organization?” the top initiative was business intelligence and reporting (72% of respondents). When asked what the purpose of Data Management initiatives was, respondents were most likely to say:
- Gain insights through reporting and analytics (78%)
- Save costs and increase efficiency (68%)
- See digital transformation as a key driver for Data Management (64%)
Participants gave the first two responses at similar rates as they had in previous years, but the number of people who stressed the importance of digital transformation increased by 11 percentage points since 2019.
“Data is the foundation of digital … That’s not going away. That’s still a big driver of a lot of Data Management initiatives,” said Burbank.
Organizations that already had a strong data foundation, Burbank found, had an easier time rapidly shifting to a digital model during the COVID pandemic. She gave three examples of such organizations that her company had worked with, including:
- A behavioral health nonprofit that moved from a 98% in-person model to a telehealth model within a few weeks
- An international bank that switched from paper-based processes to a digital workflow
- An in-person events company that used the downtime in their business to develop a data-driven culture, increasing their investment in customer segmentation, online services, and customer profiles
What all these organizations had in common was that they had to work on building a data-driven culture. “Tech is hard, tech is complicated … people are even harder,” said Burbank.
The Impact of the Pandemic on Data Management
“2020 was a lot about caution, and let’s just keep to the basics,” said Burbank.
In terms of what was not a top trend in Data Management, there was less interest in self-service data prep and in big data, which was perhaps seen as the riskier, less cautious, exploratory option. But in truth, interest in almost every category went at least slightly down as businesses went into a “holding pattern,” but the smallest decreases were in areas that give insights into trusted data – namely, Data Science and Data Governance.
“People are realizing even more so that we need to have that trusted data, so people are interested in analytics and insights and trusted data to feed those insights,” said Burbank.
Business Roles Needed for a Successful Data Management Initiative
Increasingly, Data Management is a C-level concern, the survey showed. When asked, “Who is driving Data Management in your organization?” the most common answer was “chief information officer.” There was also a 10 percentage-point increase, compared to 2019, in the number of people who said “chief data officer.” Other popular options included data architect and IT manager.
“I think so many organizations, from nonprofits to banks to retail, are seeing that data is core to their business,” said Burbank. “You can’t do that in a vacuum without business people.”
Relational Databases: A Foundation of Data Architecture
While there are many technologies available, relational databases are still a core feature of Data Architecture and the most used platform. And for good reason, Burbank said. “A lot of people call me old-fashioned. There are other tools in the toolbox … but you know, you still use a hammer when you’re doing carpentry.”
Most organizations, of course, don’t stick to just one tool but instead use a multiplatform approach. When asked, “Which of the following data sources or platforms are you currently using?” 75% of respondents mentioned a relational, on-premise database, 71% said spreadsheets, 59% said packaged applications, and 51% said they were using a relational, cloud-based database.
Burbank was troubled to see such a large share of people using spreadsheets, as in her view they are not a suitable replacement for a data warehouse or an analytics platform. “They have their place, but I wouldn’t call that an enterprise data platform by any means,” she said.
Still, there are many exciting tools that can be used in addition to relational databases, including graph databases, IoT (internet of things), and NoSQL databases, and each has its own use case. “There are so many powerful options out there right now, you really do have a lot of tools in your toolbox,” said Burbank.
The Growing Importance of Data Literacy
Finally, Burbank discussed data literacy, which she noted was taking on increased importance and more business people wanted to be involved in data. To do so, they needed to understand questions like “What’s the data quality behind this?” and “How do I read this dashboard or understand even what this metric means?” and “How do we make decisions based on that?”
“I am seeing a lot more of that where people do want to roll up their sleeves, and kind of build their own reports, which makes a lot of sense because the business knows the business,” said Burbank.
Burbank noted that it’s helpful to break out different levels of data literacy through badges. For example, a “data citizen” may be someone who knows how to use reports and brings dashboards to meetings, a “data builder” is a self-service expert, and a “data expert” may build data warehouses or do advanced analytics.
Conclusion
COVID-19 led people to return to the foundations of Data Architecture: analytics, Data Quality, and using data for digital transformation. Business insights and analytics continued to be a key driver in the Data Management space, and C-suite executives and others in business roles are increasingly involved in data initiatives.
There’s also keen interest in increasing data literacy across the organization. Relational databases continue to be a widely used platform, but there are many options to choose from. As Burbank concluded her presentation, “It’s a really exciting time to be in Data Management. There are just so many tools.”
Want to learn more about DATAVERSITY’s upcoming events? Check out our current lineup of online and face-to-face conferences here.
Here is the video of the Data Architecture Online presentation:
Image used under license from Shutterstock.com