Building a data-driven culture with a data analysis framework
Creating a data-centric culture in your business requires a shift in mindset that won’t happen overnight. However, it is a goal that becomes more attainable the more it is driven by your people working towards it with passion and enthusiasm -- as opposed to something dictated to them from the top.
There are a mix of factors that can cause barriers that hamper the successful development of a data-driven culture. At the base level, the technology and infrastructure choices a company makes, and what it provides its employees to work with, can be majorly prohibitive to a healthy data culture if that technology doesn’t support people in how they do their job.
Internal politics and silos within organizations that see certain people hoarding data can also be a problem. The IT department is typically the data steward of a business so there can be friction between IT and the rest of the business based on the misunderstanding that IT is being obstructive and trying to prevent access to data. For a data-driven culture to be successful it has to rely on a company and all its employees being able to use data in everything it does. It requires harmony between IT and the rest of the business, and a shift in systems and mindset.
Another common barrier stems from skills gaps. Organizations often fail to realize they need to teach people how to understand the value of data and help to create a skilled workforce that is capable of cultivating a data-driven culture.
Data analysis frameworks
As companies struggle to overcome these barriers, one tool that can allow organizations to successfully bridge the gap between confusion and a deep understanding of the value of data and its importance in the enterprise -- is the development of an easy to follow analysis framework to democratize data.
But how do we build and implement the framework and what outcomes can we expect? Below is an overview of the four key elements:
Your framework should start with guidelines about data preparation -- "Where is the data and how do you get to it?" This will be unique to each organization since it depends on the roles and departments that exist within the business. Whatever limitations may exist will be better understood if they’re addressed transparently. For example, limitations around the accessibility of data will help you not only understand the level of data complexities but can also create questions that help you unlock unnecessarily unavailable data.
Essentially, to cultivate data democratization and a healthy data-driven culture, you can’t keep people in the dark about the data that exists. If the process is clearly communicated all unnecessary red tape can be identified and removed.
Understanding how to approach analysis is the first step to extracting value from data.
For example, if you’ve got a new data set, where do you begin? What happened? Where did it happen? How? Why? Depending on the data you have -- let’s say in this instance it’s related to health insurance -- you can start asking questions to drill down. Where are policy holders located? How far do they have to travel for treatment? What volume of patients need prescriptions? What are the costs for those prescriptions?
Essentially, as you build a catalogue of detailed questions you start to create a guide for analysis and show the importance of asking question after question until all the necessary answers have been found.
Creating visualizations and combining them in an interactive dashboard will help users to see and drill down into the detail that helps them find insights and provide a better understanding of how variables can impact outcomes.
By using effective storytelling techniques, you can create captivating, impactful data visualizations that help people to learn something new.
Your data will largely determine the type of story you can tell. If you have time series data, your story can unfold chronologically, increasing the impact of your message from one step to the next.
When you work with very granular data it may be best to provide a summary view before guiding your audience through the detail. This sets the scene and allows you to support your argument with facts along the way.
To achieve your goal of educating your audience, you can use data visualization and storytelling to bring together different components on a single page for a comprehensive discussion of your topic.
Part of analysis and visualization is exploring data in order to trigger more questions to unearth broader areas to be explored.
Finally, you need to think about how to share findings. Creating guidelines around outputs gives those working with data a systematic approach for sharing their work and insights within the wider context of data analysis and reporting in the organization. Such guidelines are important to make data and information accessible more broadly and to help you create a data-driven mentality.
Every data discovery process should yield important lessons on top of findings. Best practices and rules should guide the work of your analysts. They will likely gain a better understanding of the readiness of your organization to become more data-driven and can evaluate whether their analyses and reports are creating replication and redundancies. From here, organizations can create rules that help improve data processes, remove unnecessary efforts and strengthen a data-driven culture.
Ethics in data should also be a part of the discussion and covered in the setting of best practices at the enterprise level with clarity on what to do with the data you have and the knowledge you gather.
The value of an analytics community
In addition to a data analysis framework, communities also have a vital role to play in helping to create a data-centric culture -- particularly in addressing skills gaps.
Bringing together your data professionals in a community within your own organization allows them to work collaboratively on solving analytical problems, to find better ways to address business questions and to optimise data engineering processes. It also provides an opportunity to create informal internal training structures through which you can increase data literacy across your business. Connecting those who are new to data analysis with experienced data analysts creates teaching and learning opportunities that connect directly to your organizations data, business processes and current data challenges.
A strong community places emphasis on knowledge exchange, learning and applying new skills, as well as networking. This helps organizations break down silos, benefit from collaboration and share resources across departments and geographical boundaries.
The best communities become a place for innovation, trial and error and establish best practices. Members can test hypotheses and discuss them with others who are working on similar problems. Through collaboration new ideas can emerge and the community, with its expertise and experience, is a great place for collecting and sharing the best ways of using data.
Instilling an effective data culture
Through the combination of an effective data analysis framework; a robust data-driven culture that empowers employees through data democratization; community engagement for teaching, learning, and collaboration; and technology infrastructure investments that make data accessible -- organizations can not only become data-driven but achieve true data excellence.
Achieving buy-in across an entire organization ultimately drives real cultural change, turning data analytics from a perceived business function into a day-to-day contributor of value.
Eva Murray is Head of Business Intelligence, Exasol. After studying psychology and business at the University of Wellington, New Zealand, Eva worked as an IT Consultant for Deloitte before moving to Sydney, Australia, as a Senior Solution Designer for Commonwealth Bank. Subsequently, she became a Tableau Consultant and Trainer for Tridant.