As we continue to push headlong into more data-driven business models, teams are increasingly expected to become analytic decision-makers. If your team has not historically had a data analytic culture, it can be challenging for them to make the transition to embrace an analytic mindset more fully. In this post, you’ll learn eight (8) ways you can start using today to transform your team culture to data-driven decision-making.
Cultural Change Begins with Leadership
Transforming your team culture to embrace a data-driven approach is not something that happens overnight. You need to instill new habits that become the bedrock of how your team approaches problem-solving and makes decisions. It will take time to develop and hone those new habits with your team. The motivation to develop this new culture must begin with leadership.
As with many cultural shifts, there may be some push back from staff who are comfortable with the old way of doing business. They know their roles and what the expectations were. Some of these staff may perceive a more data-driven approach as additional work they need to perform, or as leadership pushing off decision-making responsibilities to other staff.
Combating push back requires that you promote a firm commitment to the new approach with your team and enforce the practices that will develop those new habits. Providing staff with the latitude to exercise their curiosity more fully and engage in problem-solving exercises will instill a greater sense of ownership over the process. Be sure to celebrate the wins as they discover new insights and guide them toward the types of questions and assumptions you want tested if they get off course. If feasible, you might create a tracking and incentive system to reward staff who embrace the new practices to solidify the new approach.
Define Questions Clearly Enough to Study Them
To transform your team into data analytic gurus, you’ll need to start at the beginning of your process with the questions you ask. Eventually, you need to translate and tie every analytic question to the data to assess the answer. The best way to make this process easier to accomplish is to be very clear and specific in the questions you ask.
As you look at your analytic to-do list, ask yourself whether you have questions with the following characteristics:
- The outcome of interest is explicitly stated
- All concepts are clearly defined
- The population of interest is explicitly stated
- The time frame the data represents is clear
- Comparisons are framed explicitly to compare X to Y
For example, it’s perfectly reasonable to ask, “Has our new on-boarding process improved employee retention?” If you want to do the analysis to answer this question, however, you’ll want to make a few things clearer. You don’t need to do this as some wordsmith question-editing exercise. Instead, you can list a set of clarifying statements to go along with this question.
In the example question above, the outcome – employee retention – is clear. The concepts of new on-boarding process and employee retention, however, need to have details added to be clear what these mean in the content of the project. The population of new hires is implied by the question, but should be stated explicitly, especially if the new on-boarding process doesn’t apply to all new hires. Finally, since your interest is in whether employee retention was improved, you need to clarify as compared to what (e.g., retention last month, last quarter, or last year)?
Routinely Ask “Can we answer this?”
As your team focuses on their data-driven approach, they will begin developing a lot of questions. You won’t have time to tackle answering all of those questions, but it is a worthwhile exercise to spend a few minutes considering whether or not you can answer the question.
Have your team brainstorm about whether the data is available for the analysis. Would any data need to be collected or obtained from an outside source? If you cannot answer the entire question right now, are there parts that you could answer if you wanted to? Work with your team to identify a list of assumptions that would need to be tested for the analysis.
By regularly engaging in this exercise, your team will become stronger in thinking through the planning phases for analytic projects. Over time, they will become more familiar with the options and possible uses for the available data and more efficient at developing new projects. Finally, if you are on the fence about the feasibility of a particular analysis, you can have the team design a small-scale proof of concept test.
Develop Supporting Evidence for Ideas & Assumptions
Even with the spread of analytics throughout our lives, a lot of business decisions are still based on educated guesses and intuition. These approaches can often lead decision-makers in bad directions, even though sometimes they are the best that can be done when solid data isn’t available. When the data are available, however, it is important for your analysts to get into the habit of developing the evidence needed to support or refute those educated guesses and intuitive decisions.
Set the standard by holding yourself to that standard and taking a little extra time to develop evidence to support the direction your gut is leading you. If you don’t already have analyses to corroborate your perspective, then target those analyses first to develop support. Be sure to set the expectation for your team as well.
To keep things in perspective, however, don’t let the perfect be the enemy of the good. Many leaders have run into analysis paralysis by trying too hard to analyze their way through every decision. Effective data-drive cultures embrace the process, but also recognize the limitations of their own data and resource constraints. You still need to make a decision; the cultural transformation should help you do that.
Identify Assumptions and Test Their Validity
As part of developing supporting evidence, consider the assumptions that your perspectives are built on. Some assumptions must be true for your decisions to lead in a good direction, while others may not be so critical. Even the most experienced business people will find that their assumptions are not always correct. By taking the time to test your assumptions you can reinforce the truths you already know and correct the biases you didn’t realize you had.
Prioritize those assumptions into those that are mission critical, and those that are nice-to-know. Work with your analytic team to test the mission critical assumptions and inform your decision-making process. If you are not able to test the exact assumption you would like, but can perform a partial test, then accept that reality and take the results knowing what their limitations are.
Analytics Needs a Seat at the Table
Organizations of every size can run into challenges when leadership holds decision-making processes in the silo of a board room. Of course, there are some decisions that need to be kept among executive management. In those cases, leadership needs to provide analysts with clear instructions on exactly what analyses they need. In less sensitive scenarios, let your analytics team have a seat at the table and you will probably get better results.
When you bring your analysts to the decision-making table, they get to see and hear more about the context of the situation, the concerns and desires of leadership, and the motivation behind the analytic requests. Because your analytic team is often closer to the data, they can see connection and caveats that are not obvious on the surface. They can provide immediate feedback to help guide the planning process and improve the usefulness of analyses leadership is requesting.
Embrace A Continuous Learning Model
Once you have set the data-driven transformation in motion, you’ll want to maintain the momentum into the future. Embracing a continuous learning model will help ensure that your staff are ready for new challenges as they evolve and provide points of engagement to keep them interested and motivated.
Your continuous learning model should seek to identify and make incremental improvements rather than sweeping changes. Attempting to make large, sweeping changes increases the chance of your team not achieving the goal, experiencing increased stress, and weakening ongoing work processes. By focusing on making consistent incremental changes, you set your team up for making consistent wins without overwhelming them with new work.
As you build your continuous learning model, stay focused on your goal to improve the team’s ability to make solid data-driven decisions. Your analysts may want training on the latest artificial intelligence models, but that may not make sense in your organization. To manage the path of continuous learning, seek out the challenges your analytic team is facing and help them find solutions to become better. Explore the potential use cases for each improvement to prioritize those that need to be addressed quickly and those that can wait for a later time.
Start Cleaning and Fixing Your Data House
Transforming your team to have a more data-driven culture goes beyond the mindset changes discussed above. The best data-driven approaches will fall flat if your data infrastructure and quality are substandard. Therefore, it’s best to also start cleaning and fixing your data house to allow your cultural improvements to flourish.
Make the improvement in data infrastructure and quality a team effort. Empower your staff to identify challenges and raise concerns about how data are collected, stored, and analyzed. Work with the team to clarify the root cause of problems and develop solutions.
Aim for improving your data quality throughout the organization. Identify the data owners and systems that your analytic team needs to work with and collaborate to provide access to data while maintaining any security requirements.
Conclusion
You don’t need to be a data scientist to promote a data-driven culture. You don’t even need to be an analyst or programmer. By focusing your team on building the habits that promote asking good questions and using data to answer those questions and test assumptions, you can harness the power of their natural curiosity. Making a complete cultural transformation takes time – think months, not days or weeks. The benefits from better decision-making, more efficient analyses, and a more engaged staff are well worth the effort.
