Creating an Analytics Culture, Part II: Top-Down and Bottom-Up Traits

Read - Creating an Analytics Culture, Part I: 12 Characteristics

How do you know when your organization has an analytics culture? In the previous article, I defined 12 characteristics of an analytics culture. I grouped these traits in four categories: top-down, bottom-up, and sideways in (left and right). (See figure 1.)  

Figure 1. Twelve Traits of an Analytics Culture

At the intersection of these traits sits a senior-level data leader with broad operational responsibilities who aligns all facets of a data analytics program. The data leader can use the 12 characteristics as an action-oriented roadmap for building the necessary capabilities and relationships to create a data-driven organization.

This article focuses on top-down and bottom-up traits. Top-down refers to the words and actions of top executives, while bottom-up focuses on how knowledge workers use information to do their jobs. Knowledge workers are both power users, such as data analysts and scientists, and casual users, such as data consumers and explorers. (See “How to Succeed with Self-Service Analytics, Part II: Know Thy Customer.”)

Top-Down Traits

Executives set the tone and cadence for an organization. What they say and do trickles down and influences the attitudes and behavior of every worker. Employees take their cues from top executives, both in good faith to support the corporate mission and to secure their positions and livelihood. No one wants to “step out of line” and alienate a boss or executive who has the power to eliminate their job.

Data and analytics are not easy. Going from intuition- to data-driven decisions requires time, talent, and investment. Most importantly, it requires that people change the way they work. Such upheaval doesn’t happen without strong executive leadership and support.

Executive Commitment. So, the first and most important indicator of an analytics culture is executive commitment to data. In organizations with analytics cultures, top executives understand the importance of data and ensure their organizations manage it carefully. This commitment is usually manifested in the form of an “executive mandate”, ideally issued by the CEO.

When the CEO writes a memo to employees about the importance of managing data assets or insists on a common set of metrics for measuring company performance, things change, usually immediately. Grassroots initiatives around data quality, data governance, performance management, and master data management that struggle to gain traction now gain extra attention and funding. Business unit managers that resist centralization or standardization efforts fall in line and take an active part in alignment activities.

Executive Example. But words only go so far. What an executive does has a far greater impact than what he or she says. Every parent knows that children follow their example more than heed their words. The same is true in organizations. An executive who issues a mandate for standard data and reports but then makes decisions with spreadmarts sends a loud and clear signal to employees that it’s ok to create data silos and spreadmarts and ignore or resist corporate calls for standardization, consolidation, and governance.

Conversely, executives who use standard data and reports to make decisions create a ripple effect throughout an organization. Managers who report to data-driven executives know they need to study the executive dashboard so they aren’t asked questions they can’t answer. The fear of embarrassment and possible rebuke forces managers to become more data-driven themselves so they can defend their plans and justify their actions. This process trickles down from managers to subordinates throughout the organization.

Executive Decisions. Another factor is how executives use data to make decisions. Research shows that the best decisions balance experience and data. Executives who understand the need to use data to validate intuition and intuition to validate data create organizational habits that result in better decisions. This is why some people now talk about “data-informed” decisions rather than “data-driven” ones. Data without historical experience filtered by humans can also lead to poor decisions.

Similarly, If executives don’t review dashboard and scorecard findings in regular operational or strategic review meetings, lower-level managers and their direct reports learn to ignore requests for performance metrics from the executive suite. They might also stop using data when making their own decisions and plans. Executives set the tone for how the entire organization uses and manages data.

Bottom-Up Traits 

Another way to tell whether an organization has an analytics culture is to observe how knowledge workers view and interact with data. At data-driven organizations, knowledge workers receive the tools and training to turn data into insights and actions and conform with processes needed to govern data.

Data Stewardship. In an analytics culture, business users trust data and actively govern it. Data owners and data stewards are responsible for managing the meaning and quality of data in their domains. Data stewards continually check the accuracy of critical data elements within standard reports and dashboards and work with data administrators on the data team to fix errors. They also recommend changes to critical data elements and escalate issues to a data governance council.

In an analytics culture, every business user takes responsibility for governing data, not just stewards and owners. For example, business unit heads follow the lead of corporate and create a standard set of reports for their departments that address most questions staff have on a daily basis. Data analysts transfer ad hoc reports to a development team who converts them into production reports. Users generate tickets when they see errors in reports or alert a data steward or data administrator to the issue.

Data Empowerment. In an analytics culture, knowledge workers have self-service tools that empower them to find, query, manipulate, visualize, and share data without IT assistance. Data analysts take responsibility for building the majority of local reports and dashboards. They work closely with the data team who primary focus now is to manage enterprise data, build complex, cross-functional applications, and facilitate self-service analytics.

In an analytics culture, organizations empower business users with the type of self-service best suited to their role and requirements. For example, data consumers want interactive dashboards and views they can save as snapshots. Data explorers want to modify existing reports and dashboards via a semantic layer that lets them add fields, dimensions, calculations and so on. Data analysts want to create custom data sets and dashboards that enable them to explore and analyze data. Data scientists want to model historical data using statistics, machine learning and artificial intelligence.

Data Literacy. Finally, empowered users are informed users. An analytics culture has an immersive training program designed to help business users across the spectrum better interpret and harvest data. These organizations boast a training program that teaches business users how to use data and analytics tools, interpret corporate data, and take appropriate actions.

Analytic cultures also promote informal training channels. Most foster a community of interest (CoI) among data analysts who meet regularly, both in person and online, to share tips, tricks, and techniques. Data scientists mentor newcomers to the team and work collaboratively to solve problems. The data team holds office hours and labs to help power users sort out technical issues. In turn, power users provide ad hoc training and coaching to casual users who need help using self-service tools.

Conclusion

You can recognize an analytics culture first and foremost by observing the behavior of top executives. In a strong analytics culture, top executives make public commitments to advance data and analytics, set an example by using standard reports, and balance data and intuition when making decisions. Secondly, these organizations train employees how to use data analytics tools, interpret data, and take actions. They also establish governance processes and roles so business users take an active role in governing data.

The final article in this series examines the traits that departments and data teams exhibit in organizations with an analytics culture.

Read - Creating an Analytics Culture, Part III: Sideways-In Traits

Wayne Eckerson

Wayne Eckerson is an internationally recognized thought leader in the business intelligence and analytics field. He is a sought-after consultant and noted speaker who thinks critically, writes clearly and presents...

More About Wayne Eckerson