Part two of our data mini-series explores how data can have unintended consequences in work and life and introduces the strategies we recommend for overcoming the challenges in assessing data sources and responsibly applying insights from data in your decision-making process.
Lessons from Amazon …. and Me
As data use has become pervasive and unavoidable in society, it can become increasingly difficult to separate the good, from the bad and the ugly. In some cases, data must be dead-on accurate. For example, a smart car with autonomous driving features must receive and interpret data to successfully make decisions that could be life or death. This is an example of good data processing in action.
Bad data processing can lead to bad decision-making. An example is when Amazon received severe backlash after reports of employees using bottles to go to the bathroom at their workstations because of limited breaks caused by data-driven decisions to increase productivity. Amazon missed an opportunity to delve into understanding the story behind the data. Rather, it set targets and metrics for its employees purely based on numbers. Ponder an example of teachers who are penalized for not meeting goals for virtual student learning. Simply relying on numbers without context, penalties were issued to these educators without considering the root causes. Specifically, students in low-income households could not afford devices. Even when they received donated devices, they did not have adequate internet access to utilize those devices. Looking beyond the numbers can lead to improved decision-making that targets the root of the issue. There are many benefits to data, but we must be aware of the risks if we don’t critically evaluate how data is collected, processed, and utilized for higher-level decisions.
A personal example showcasing the importance of understanding the story behind the data was during the early emergence of COVID. Patients could not be seen in-person and healthcare was converted to virtual visits. We developed a patient portal to grant access to health information data online (an amazing feat in itself). As the institution circulated goal targets for enrollment in this online portal, I noticed my team and division were below targets. While discussions focused on pushing the goal and signing up more families to use this helpful resource, I decided to step back and critically evaluate the data for my team. Rather than asking what we could do to push the numbers higher to meet pre-set targets, I decided to question the data and ask why the numbers were so low in the first place.
Data Talks – How to Listen and Question Data for Better Decision Making
In this ongoing tale, I did what I am asking you to do – ask questions of your data. As a leader, it is important that when you are presented with data, you automatically flex your thinking and ask yourself a few questions – How was this data captured? How is this report generated? What is it supposed to measure? Is the data measuring what we really want to measure? Are we asking the right questions to capture this information? If not, why not? Always ask – what is the story behind this data?
So, I did just that. Rather than bulldozing our way to hit just a blind target, I questioned, “What is the real story this data is trying to convey?” Upon further inquiry, the story behind the data became evident. The “input” was revealed, and I was able to see why the “output” was falling short of targets. I discovered that non-English speaking patients had barriers to signing up and using the online portal because interpreter services were not available. This led to frustration from families. Ultimately, they couldn’t utilize this excellent resource that the organization had built to help them. I connected with the tech team at my hospital and discussed these findings with them. I advocated that rather than pushing a telehealth goal for providers and teams, the infrastructure in place needed to be aligned in order for providers and teams to meet and exceed those goals. Thus, the organizational focus shifted from driving up the numbers, to creating a system that actually led to the improvement in care by decreasing barriers to health care access. All of this would not have happened if I hadn’t challenged myself and my team to ask questions about the data and investigate the story behind the data. Sure, we could have signed up more people for the service and we would have hit the target number, but we would have left many families behind that had barriers to access.
The Way Forward – Data Literacy
By now, we have seen why meaningful data is very critical to successful decision making; but how do you define and capture meaningful data? Do your data management practices emphasize measuring numbers and overvaluing systems, or are you looking beyond the data to truly understand what is being measured and the story being told? Organizations can risk losing talent, or face challenges with recruitment and retention, if metrics alone are relied on for strategic decision making. Individuals and organizations should strive to improve the data literacy of their teams. Gartner defines data literacy as one’s ability to “read, understand, create, and communicate data as information.” While there are many experts on this subject, I’ve collected a few tips below to help you in managing data and optimizing your decision making in a metrics-driven world.
Selection & Governance: Starting with the right inputs may be the most important consideration. Data of high quality is considered trustworthy, accurate, relevant, and complete. Governance is the process of formally managing data assets to maintain quality and security. Not all data is appropriate or relevant, nor should it be accessible to all audiences. Taking care to be intentional when selecting data and ensuring that this data is controlled to preserve quality will be critical for collecting meaningful data.
Transformation & Analysis: Tim Stobierski, Harvard Business School Online, describes the key data literacy skills and concepts for business. Transforming your data prepares it for analysis. Often, raw data must be “wrangled” or “cleaned” in order to be used. Stobierski goes on to describe the four types of analysis. Descriptive analysis explores “what has happened.” Diagnostic analysis evolves this understanding by diagnosing the why behind the what. Predictive analysis is the process through which predictions are made about “what might happen.” Finally, prescriptive analysis is the decision-making process to achieve a “desired outcome.” Once you have selected and transformed high quality data, coordinating strategic analysis to discern insights and conclusions will ensure that data is being integrated in your business responsibly.
Visualization & Reporting: The process of visualization requires that you have processed the data such that others may easily consume this as information. There are many ways to visualize and share the output of your data collection and analysis. Consider who your audience is before tailoring your delivery. Understanding how and why this data impacts them will ensure that you deliver richer insights.
Lastly, remember that data always tells a story. It is your responsibility to listen carefully and accurately to what it is saying. Your ability as a leader and manager to master data interpretation are indispensable for setting accurate objectives and meaningful key performance indicators for you and your team. Here’s to improving our data literacy and mastering data management to effect change as we sort through the good, the bad, and the ugly.