Welcome to the Collibra Blog, where CDOs, data stewards, and data citizens go to learn about true data governance.


Winning with Data (and How Data Governance Can Help)

Share: Share on FacebookShare on Google+Tweet about this on TwitterShare on LinkedInEmail this to someone


In the book Winning with Data: Transform Your Culture, Empower Your People, and Shape Your Future, authors Tomasz Tunguz and Frank Bien explore the culture changes data brings to the business. As someone who talks about data governance every day, I found this book very relevant to the discussions I have with customers and prospects every day.

The creation of a data-driven culture within an organization is key to the success of today’s companies. To enable this culture, management and employees expect answers to their data questions in a near real-time fashion. And while this demand of ‘instant data’ becomes the expectation rather than the exception, many organizations struggle with it. At their core, the best data-driven companies manage to operationalize their data to drive business decisions, optimize their market, drive growth, and predict where the company is going. Modern data infrastructure is a necessity, but is insufficient to convert a company into a data-driven market leader. The journey to become a real data-driven company requires a difficult change management exercise to transform the company culture and make data part of every discussion and decision.

Data breadlines, the typical track one needs to follow in the organization to request answers from the various data sources, are only serviced by a handful of data analysts. Therefore, every request goes through a system of queues and prioritization. As a result, teams have to wait extensive amounts of time to even ask their question to the analysts in the first place.

Once they have reached the front of the queue, ‘data obscurity’ prevents business users and analysts from finding a common language to define those questions and understand where the data providing answers exists in the organization.  To make things worse, the above issues are often ‘solved’ by employees who create their own data stores and shadow analyst teams. This leads to ‘data fragmentation’ (data being spread across the organization). With those different data sources supposedly carrying the same data and feeding similar reports, meetings can quickly escalate to ‘data brawls’ where employees endlessly discuss which report or data stores contains the correct information. The conclusion is clear: “without a universal lexicon, confusion is inevitable and conflict unavoidable.”

Therefore, a key factor in creating the ability to operationalize the data is a data dictionary which contains a canonical definition of each metric and where the supporting data can be found. This enables productive and incisive conversations about data across teams, bolstering or refuting argument and accelerating decisions. In addition, true data-driven companies create equations that describe their business and capture all contributing parts to the revenue. Within these equations, the variables are the things to measure, the very same metrics that can be found within the data dictionary.

Traditionally, decisions were made by lots of opinionated debate, followed by the most senior person selecting a path forward. But true data-driven decision-makers acknowledge that great ideas do not always come from the most experienced or most senior member of the team. Great ideas can originate from everywhere, both internal as well as external to the company, and data is often the best tool to ensure those ideas rise the ranks.

A cultural shift within data-driven companies enables people to use data in order to avoid conversations to become political or emotional and no longer focus on the key issue at hand. In addition, it slays the HIPPO (highest paid person’s opinion) as the determining factor of the direction of a project, team, or company. When we make decisions with data, we decide collectively because we’re all evaluating the facts and interpreting them as a team. This aids collaboration within the team and helps collective ownership of the decision and ultimately helps shifts the culture.

This culture shift starts with curiosity. Empowered by data, curious employees can create hypotheses, design experiments, validate ideas, and change the company’s direction based on them. However, as curiosity cannot be learned, it’s of great importance for organizations to hire based on cultural fit to the company. A clear definition of the culture is an essential part of building a successful recruiting process and hiring the fitting profiles.

True data-driven companies want every employee to discover new information, understand it within the context of their existing frameworks, determine what assumptions are no longer true, and then evolve the way they think about the world to move things forward. Business intelligence has always been an important part of the puzzle and a long road has been covered since analyzing data using the simple queries on the mainframes. But with extreme data collection being new normal, and market leaders in the business intelligence space struggling to catch up, organizations demand retooling of the analytics that access the vast amounts of data and provide the answers the company is looking for.

To solve the bottleneck in the data breadlines, a successful strategy that many organizations employ is decentralizing data analysts to be part of the different business teams. Over time, this will enable the data analysts to focus on the most difficult challenges.

This change has to be complimented educating employees in data literacy. This includes teaching the right tools and mind-set that will enable them to use the data to their, and the company’s, advantage. Furthermore, employees have to be educated on what data and tools are available and how to analyze and visualize the result. An essential part of this educational process is making employees aware of the different types of biases that exist within data and how to avoid them.

So, where do we go from here? Historically, organizations used data to create hindsight by analyzing what and why something happened using descriptive and diagnostic analysis. However, by creating a data-driven culture, organizations are able to provide insight by showing what will happen by using predictive analysis and going forward, organizations will continue their data-journey and answer questions on how we can make it happen by using prescriptive analysis to create foresight. However, it is important to realize that regardless the type of analysis an organization performs, actionability is a key attribute of useful data. Analyzing data for the sake of it is simply a waste of time.

To sum it all up, data governance is able to build this ‘universal lexicon,’ a single platform where we can find, understand and trust the information in the organization. The platform keeps track of all data requests, contains a business glossary linked to a data-dictionary to avoid ‘data obscurity,’ and keeps track of ‘data fragmentation’. In addition, it provides a place where resolutions to ‘data brawls’ can be found. All of the above are essential factors to make data breadlines more effective and less painful, and to support the journey to become data driven.

Benedict is a Sales Engineer focused on the Financial Services industry. Before joining Collibra, he worked for Wolters Kluwer Financial Services as principal consultant implementing Finance, Risk and Regulatory compliance solutions worldwide.

Comments are closed.