When we talk about data-driven cultures, what do we mean? What are the individual and organizational characteristics we look to cultivate? In this series of blogs, we will share three key strategies relating to people, process, and organizational change that lead to successful data-driven cultures.
Our prior post on data-driven cultures discussed how adopting the right mindsets can improve your decisions by combating the cognitive biases that can cloud your judgment. In this post, we examine why you need to systematize your decision-making processes to reinforce the behaviors found in data-driven cultures.
Examining your Decision Processes
As we noted, data-driven cultures are obsessed with an open-minded pursuit of the truth, supported by rigorous analysis. The combination of these two factors determines the quality of decisions. To accomplish this, you need to channel your obsession into creating a systematic process that can be learned, practiced, and improved daily. In other words, your daily behaviors and actions are guided by a clearly defined decision process.
There are three reasons for doing this:
Before we go any further, let’s define a decision process and how it differs from a business process. A decision process is a set of discrete steps individuals or teams use to evaluate and analyze information to create insights that fuel decisions. At a high level, it involves locating and evaluating evidence, structuring it for analysis, and making decisions based on this analysis.
Every decision involves predicting the future and, ideally, keeping score as to the accuracy of your predictions. Decision processes usually are embedded in a broader business process, with the distinction that the output of a decision process is an agreement to take some action. The output of a business process is typically a tangible thing (e.g., a product or service) resulting from many cumulative decisions.
Examples of decision processes include a consumer goods manufacturer engaging in demand planning to inform production schedules. Or the finance function of a fast-growing tech firm projecting cash flow and profitability to ensure hiring new staff doesn’t outpace financial realities.
A complete discussion of all the decision types is beyond the scope of this post. Still, a critical takeaway is that the mistake most leaders make is underestimating the complexity of decisions, which translates to overlooking opportunities for improvement. Furthermore, improvement doesn’t mean a single, significant change (which can be enticing but unattainable) but incremental progress over time and dozens of decisions.
DPI: Change and Continuous Improvement
To improve decision processes, teams must first understand how the process currently works, make assumptions about what might improve it, align stakeholders around the goals for improvement, and embed the new behaviors in the decision processes. This is straightforward -- even intuitive-- but is rarely done inside organizations. In their book, The Knowing-Doing Gap, two Stanford professors may have found the reason people too often know what to do but still do not follow through:
“…the answer to the knowing-doing problem is deceptively simple: Embed more of the process of acquiring new knowledge in the actual doing of the task.”
In other words, continuous improvement of your decision process is not an intellectual exercise; it’s a formal practice. We call this formal practice Decision Process Improvement or DPI. At its core, DPI is about improving operational performance by increasing the speed and quality of decisions. Let’s dive in to see how it works.
How DPI Works
Using DPI, teams start with a decision process (e.g., loan processing, sales or financial forecasting) and work backward, looking at all the inputs and outputs that informed the decision and identifying areas for improvement that will impact the speed and quality of decision-making.
The key elements of DPI include:
Common Decision Process Pitfalls
Over the years, we’ve worked on projects across industries building advanced analytics solutions for a wide variety of challenges, including fraud detection, manufacturing demand forecasting, pricing optimization, product defect rates, sales forecasting, and loan portfolio risk. While each of these use cases has slightly different underlying decision processes, we’ve seen a consistent pattern of decision process pitfalls.
The following are some of the contributing factors, organized by high-level decision process phases, that we’ve found lead to slower, lower quality, and more resource-intensive decisions.
Major Decision Process Phases and Common Pitfalls
Gathering and organizing data
Annie Duke, author, consultant, and accomplished professional poker player, attributes her success over the years in making rapid, high-stakes decisions with real money on the line to developing a high-quality decision process.
"Why is it so important to have a high-quality decision process? Because there are only two things that determine how your life turns out: luck and the quality of your decisions. You have control over only one of those two things. What you do have some control over, what you can improve, is the quality of your decisions."
Once you raise the visibility of your decision processes and commit to continuous decision process improvement, you will get closer to the data-driven culture you need to make faster, better decisions.
Editor's note: this post, co-authored by Rick Hinton and Lisa Targonski, originally appeared on the Elder Research blog
Read Part 3: Using the CALM Method to Sustain Change