In one recent week, I heard about the need for increased “data literacy” from executives at a major insurance company, leaders at a large consumer packaged goods manufacturer, and senior administrators at a preeminent research university. All these leaders understand the need for a comprehensive data strategy and want to create one for their organizations. Yet, they are finding it difficult to effectively communicating their strategy to employees who are not used to working in a data-driven environment, and who don’t understand how to implement that strategy in their day-to-day work.
Data Scientists don’t give much conscious thought to data literacy; we’re native speakers. So, what is data literacy? Why should it matter to managers and executives? How can you or your employees acquire greater data literacy?
What is Data Literacy?
Alberta Education defines general literacy as the ability, confidence and willingness to engage with language to acquire, construct, and communicate meaning in all aspects of daily living. Literacy is inextricably tied to information and communication.
Literacy does not imply mastery. Being literate in the art of writing doesn’t make one a poet. Being financially literate doesn’t make one able (or want) to trade equities. Rather, literacy encompasses confidence, willingness, and understanding.
I argue that data literacy is the ability, confidence, and willingness to engage with data to:
- Acquire information
- Construct or review data analyses
- Communicate insights to inform decisions
Why Does Data Literacy Matter?
Sifting Value from the Hype
There is a lot of excitement around Data Science, Artificial Intelligence, and Machine Learning. To institutionalize a data-driven culture, it is important for executives and employees to be familiar with the state of the art in these fields, and relate them to their unique business’ needs and challenges.
Organizational leaders who are data literate recognize opportunities to derive value from their data while avoiding the hype. They can define and effectively communicate the business value of Data Science or Machine Learning initiatives from the onset (not as an afterthought once an initiative is complete).
Selecting Valuable and Actionable Metrics
Which is more important: increasing sales or profitability? Pursuing either goal seems to reinforce the other, but the reality can be subtle. Focusing on increased sales may incentivize decisions that erode profitability. For example, applying a discount will result in less profit per sale, even though total sales increase. Conversely, raising prices will increase the profitability of each sale, but will likely result in fewer sales overall. It is vital to select the right metric by which to measure success when using data analysis.
Data-driven managers understand their operational data and the information they convey about the health of the business. They know the values that matter and how to manage them to drive future success. Choosing the right metrics matters because metrics establish the goalposts by which we determine the success or failure of the initiatives.
Understanding the Data Science Lifecycle
If you’ve spent a few years in your current organization, you probably understand your business processes and how to generate a valuable product or service for your customers. Data science has its own lifecycle, the major stages of which are:
- Define the problem
- Gather relevant data
- Identify important variables
- Create and validate a model
- Deploy the model into a business process
- Monitor results and refresh models
These stages overlap and intersect with existing business workflows.
Data literate executives understand the challenges inherent in each of these steps and appreciate why each one is necessary. Executives should understand this lifecycle and assess if business goals are supported by the analytics project plan and provide guidance to ensure that the Data Scientists work toward an outcome that is valuable to the business.
Making Well-Informed Decisions
Business leaders want to make informed decisions to support their goals and grow their business. They have a variety of data sources at their disposal to assess the value and risks of different options. However, data often reveal limitations or biases in our thinking and it is important to ensure that the data accurately captures the behavior of a representative sample of the population (e.g. customers or other stakeholders) about whom insights are required. Data literate leaders understand these data challenges (and models produced from data) and can help mitigate them to support effective decision making.
Ensuring the Ethical (and Legal) Use of Data
It is tempting to believe that because models are analytical that they are unbiased or above human influence. Some industries (e.g., banking) have strict rules about what information can be used for decisions. Others, like medical device qualification, have explicit requirements for transparency regarding algorithmic methods. Data literacy helps avoid embarrassing ethical lapses or costly legal entanglements.
How Can Executives Become More Data Literate?
Read, Read, Read
There is a wealth of articles that dive deeply into the facets of Data Literacy. Three references I recommend:
- Mining Your Own Business — Our CEO, Gerhard Pilcher, and our Chief Operating Officer, Jeff Deal, provide excellent advice to executives on how to lead data initiatives based on Elder Research’s nearly 25 years of experience with Data Science. (Download Chapter 3 – Leading a Data Analytics Initiative)
- HBR’s Data Analytics Basics for Managers is an excellent summary compilation of their highest quality articles on Data Analytics and Data Science.
- Tom Davenport’s and Jeanne Harris’ Competing on Analytics is an early and influential work in this space, and remains highly readable and relevant.
You don’t need to complete a Coursera certification in R or Python to become data literate. For example, the Elder Research half-day Executive Strategy Session has helped executives across industries develop or refine an analytics strategy and project roadmap to support their organizational priorities. Additionally, we offer a range of analytics training courses targeting analytics practitioners and managers, from half-day, focused skills sessions to 3-day, intensive data science proficiency courses.
Seek Out Trusted Data Advisors
Some questions require expert, outside advice. Even the literate need to consult masters on important matters. Consider hiring a specialized, data analytics consultant. Elder Research has helped businesses, government entities, non-profits, and universities find value in their data for longer than almost anyone in the world. Once we understand your goals and priorities, we draw on decades of experience to build and deploy models that deliver results and inform process stakeholders. We also work with in-house analytics teams to provide third-party validation of their models and methods and build an analytic center of excellence.
Build Relationships with Your Data Team
If you’ve recently hired new Data Scientists or Data Engineers, it’s important to establish a collaborative relationship between the technical team and business process owners as you plan your data strategy. They are the cross-functional team that will help you define the horizon of possibilities, so be open to their feedback. If an idea seems great to you, but they seem tentative, ask them why. While you may have greater visibility across the business, they will be more familiar with the nuances and limitations of your data. Working together will ensure their efforts align with overall business strategy. You may want to mentor them in how to communicate the business value from their findings to key stakeholders to increase adoption of their results.
Data literacy is the ability, confidence, and willingness to engage with data to acquire information, construct or review analyses, and communicate insights to inform decisions. It is distinct from proficiency. The data literate are not quantitative prodigies, but are familiar with the methods used to create models from data, apply a critical eye to any insights derived from data, and communicate those findings to inform decision making. Being data literate is essential for executives and managers to enact a data strategy that will succeed in their organizations. And, it is achievable through self-learning, training, expert consultation, and building relationships.