TDWI | Training & Research | Business Intelligence, Analytics, Big Data, Data Warehousing

Executive Perspective: A Closer Look at Data Literacy

How is the concept of data literacy evolving? Roman Stanek, CEO and founder of GoodData, discusses how organizations need to move beyond last year's model to help users derive more business value from data.

• By James E. Powell
• March 1, 2021

Upside: Data literacy was a big buzzword in 2020. How do you define data literacy and why is data literacy crucial for organizations?

For Further Reading: CEO Perspective: Moving Data Analytics from Cost Center to Revenue Generator Why Data Literacy is Critical to Higher Satisfaction and Productivity Why Analysts Must Move Beyond Data Literacy and Focus on Business

Roman Stanek: Data literacy is simply being able to understand and interpret data effectively. However, it's an interesting question because I'd argue that literacy is the wrong buzzword. Companies need to be thinking about deriving action from data. Instead of putting the onus on the data consumers to understand complex data sets, the responsibility should be squarely on those delivering the data to make it more digestible and prescriptive from the start.

How does an organization measure data literacy?

Businesses can begin to measure data literacy in a simple way: how many people in their company/ecosystem are effectively using data to make decisions within their daily workflows?

Historically, data analytics as a business function has been relegated to specific, often technical, departments within an organization when in reality it should permeate throughout. When the data is presented in a digestible and consistent way across an organization, that's when businesses start to see significant ROI -- which is a more precise way to measure literacy.

How can companies increase the data literacy of their staff?

For too long the responsibility of interpreting data has been on the end-user -- and that has led to low data literacy. We actually need to shift accountability entirely and stop seeing low data literacy as user error. Instead, data needs to be pre-interpreted to boost data literacy across any organization.

It really starts with providing better-designed, data-driven applications for their employees. When the complexity has been interpreted and data is instead delivered in the form of intuitive assets, it is far easier to manipulate, interpret, and, most important, act on the implications of data.

Do you believe the onus of literacy should be on those delivering the data or those interpreting the data? Why?

I think it's imperative the onus is on those delivering the data. Instead of giving people a copy of a CSV (comma-separated values) file and a data visualization application and expecting users to pull insights out of raw data, businesses should be delivering a user-centric, designed experience. Technical teams should still know how to understand data, but that skill shouldn't be required of everyone in an organization -- and certainly should not be the barrier to actually acting upon data insights.

It's key for analysts to be able to access the data they need. What is the difference between accessibility and literacy and why does that matter?

Even though data accessibility and literacy are different, they go hand in hand. Accessibility is all about giving the highest number of people across an organization access to data, whereas data literacy is about interpreting said data.

The goal is to create a system that increases both accessibility and literacy in tandem so that data-driven decision making is as easy as possible. With better back-end applications that replace the ambiguity of CSV files and error-prone formulas, companies can seamlessly distribute data and present pre-interpreted and digestible data for greater efficiency and enablement. The end result is a new data culture within their organization -- one centered on ease-of-use, informed decision making, consistency, and trust.

When we spoke a few months ago, you discussed how companies can monetize their data, essentially taking data analytics from a cost center to a revenue generator. Has this happened to the extent you expected?

Yes -- and to be clear, by data monetization I do not mean that companies should be selling data. I mean that companies can create new revenue streams from the insights gleaned from data, not the data itself.

Since we last spoke, some companies have begun monetizing their data. By increasing insights consumption across their organization, they've been able to derive real ROI from data-driven decision making. However, there is a real opportunity for more businesses to get on board. Especially as the pandemic persists, it has become even more imperative that companies double down on new opportunities that drive revenue. I firmly believe that data is the new revenue generation tool they've been looking for.

Do you believe that more companies will achieve this in 2021? What's shifting in the industry to make this true?

Yes. We've already seen major changes in the data world with Snowflake capturing an entire cloud data market. I believe Snowflake will gain a true set of competitors, which will change the data landscape as we know it. Rather than slow and cumbersome data warehouses, the world's data will be stored into standardized cloud storage, which will redefine how data is managed in every company and realign the data value chain.

We're also seeing a push around insights consumption as the next step. When businesses can effectively make data-driven decisions across functions, that is where we'll see massive growth.

What do you predict will be the biggest change in the data landscape over the next few years that will impact an enterprise's level of data literacy?

For years, analysts have limited the definition of business intelligence to only data exploration tools that specialize in dashboards, but there is a big difference between data visualization and the back-end creation of a data engine. Today, when data analysts write formulas based on CSV files, for example, they could come up with thousands of different versions of the truth. However, when you have complex indexing happening in the background of a data-driven application, and you've translated the complexity into everyday business terms, you're left with a more intuitive and accurate front-end experience.

Thanks to years of pigeonholing BI, we are left with a term that only stands for spreadsheets and data dashboards. Instead, we need a new definition of enterprise data literacy -- one based on insights consumption and the ability to make truly data-driven decisions.

What role do predictive analytics, AI, or machine learning, play in data literacy?

We're at an interesting juncture where a collective move to the cloud is escalating growth and modernization in the industry similar to what Y2K did for software. This is the biggest opportunity since Y2K to reorganize data analytics and reinvent the norms of how businesses use data, breaking away from the muddled, fragmented approach that's pervasive today.

Predictive analytics and AI/machine learning will ultimately help create even more accurate and streamlined modeling to scale data monetization for the decade to come.

[Editor's Note: Roman Stanek is the CEO of GoodData, a company he founded to disrupt the business intelligence space and help organizations monetize big data. With over two decades of leadership, Roman has become a strong voice in the analytics industry, pushing the boundaries of how companies use data insights to move forward. Prior to GoodData, Roman served as founder and CEO of two startups, NetBeans and Systinet, which were successfully sold to Sun Microsystems and Mercury Interactive (later acquired by HP software) respectively. You can reach Mr. Stanek via LinkedIn.]