Today’s organizational challenge: From gut feeling to data-driven decision making
Most organizations today recognize that analytics has great potential to create a competitive advantage. However, many organizations still struggle to reap the benefits of fact based decision making and keep relying heavily on gut-feel and consensus decision making.
Jorg Schalekamp, Marjolein Vlaming & Bart Manintveld - 5 October 2015
Creating value from analytics is not a technological or data challenge anymore, it is an organizational challenge. Recent research indicates that most organizations aren’t ready to step up to the challenge. Symptoms of failure are for example too little stakeholder engagement, unclear (analytics) strategy, vaguely defined goals, misalignment of priorities and ambitions across business units, hidden agendas, lack of leadership commitment, and too much focus on gathering data, building models and technology solutions. In general there is too little attention for user engagement right from the start, capability building and making analytical tools simple and engaging.
Sitting at the heart of these issues is a central question that organization’s need to resolve: How can we get people to change their behavior and use analytics in their day-to-day jobs and rely on facts?
Analytics and being willing and able to change
Let’s first get a conceptual view on how change works. The two important pre-requisites of change are that people are both “willing” and “able” to use the tools provided. Though the result of analytics might seem a technological solution (e.g. dashboards), the real successful outcome would be the use of that solution in a users’ day-to-day job. A good technological solution is a pre-condition for success but, right from the start, the human aspect is important.
The “willing” and “able” prerequisites should be targeted with different interventions. Several scientific models about the adoption of technology underline the importance of these non-technological/ human aspects.
For instance if people aren’t able to change, the solution can be found in building the right capabilities e.g. by training, coaching or ensuring effective organizational design. On the other hand, the “willingness” of people asks for different interventions. Are your employees committed to carrying out specific actions? Do they feel they belong to the part of the organization that needs to change? Do they share a common understanding of how things are done? This can be applicable in the case of analytics as an itinerary: “People clearly understand how to reach the goal of the analytic solution, they just don’t feel like walking that path”. Think about what questions your employees will ask themselves when an analytical tool can outperform their decision making skills?
Four actions to take to facilitate data-driven decision making
We present four practical actions organizations can take to effectively support employees in embracing a more fact-based decision making style. It is not an exhaustive listing of all possible actions to take to enable data-driven change, but a trigger to shift the lens that analytics is not only about data, modelling and tools. It is about people acting differently to reap the benefits.
# 1. Create ownership and understanding by involving stakeholders in the design phase
Solution design is not just about creating the best possible analytical tool or model. The word “design” suggests the use of design thinking techniques, such as stakeholder engagement, multiple iterations and prototyping. The involvement of relevant stakeholders from the start, e.g. by developing the tool together, can support you in building their ability and willingness to support the solution and in creating ownership and understanding.
# 2. Assess your organization’s change readiness
Conducting a change readiness assessment can help you understand what possible risks and opportunities exist within your organization related to the change. This can support you in defining your change approach. New data driven change readiness tools like the Change Adoption Profiler use data and techniques from marketing to provide insights into your organization’s attitude towards change and allow you to address it head on.
#3. Be aware of the impact of a provided solution on each of the involved stakeholder groups
The impact the implementation of a new tool can have on various stakeholders can differ tremendously. A change impact assessment can help you understand which groups within the employee population are impacted most by a change. This assessment can help you determine your key messages, communication plan or possible employee-specific interventions. Drawing an “employee journey” can help you understand in which step an employee is positively or negatively impacted in a specific process. For example, sales employees will use a customer segmentation differently and at different times than the people from marketing. Having a clear view on how your tool impacts their roles will support you in engaging them in the most effective way throughout the project.
# 4. Create a campaign for your analytics solution and tailor your approach to your target group(s)
“Change is a campaign, not a decision. CEOs can demand, but the people must want to act. Visions must be sold over and over”. This means communication from and to different stakeholder groups involved in the analytic solution is key. What impact would it have if a leader stimulates the use of analytics within the organization? Your “change campaign” should provide clear benefits around the change that can be tailored to specific groups. The method you use should best fit the group’s needs.
Analytics as a compass
Analytics can guide employees in two different ways – as a compass, or as an itinerary. Each way has their own challenges.
In some cases, a model can only point the user in the right direction, but not suggest how to provide the required behavioral change. Customer Segmentation can for instance separate clients into meaningful groups, but it is still up to the marketing or sales team to decide how to act on this information and translate it into targeted interventions. In this case analytics function as a compass.
Analytics as an itinerary
For binary decision making tasks where large amounts of data are involved, such as whether or not to hire a person or detect if someone is committing fraud, several studies have demonstrated that simple statistical models easily outperform even leading experts in the field*. In these cases a model makes a clear recommendation – analytics functions as an itinerary.
A famous practical example of this in recruitment is ‘Moneyball’, the story about baseball manager Billy Beane's successful attempt to assemble a winning team on a lean budget by employing analytics to acquire new players. Though the benefits of using analytics in decision making are clear, the use in recruitment is still not widespread. Thus even in the case where the appropriate action is straightforward, it can be a challenge to get people to act according to the model’s recommended action, instead of following their own insight.