Building a Culture of Analytic Excellence
Design of Experiments: Why Do I Need Data Analytics?
Researchers, developers and engineers may question whether they have the capacity to adopt data analytics into their busy working lives. When there is always more to achieve than there are hours in the day, how is it possible to become an expert in data analytics as well? Colleagues and management teams do not always see the benefits of data either, happily entrenched in decision making based on a combination of subject matter knowledge, anecdotal evidence and traditional approaches to experimentation. The resistance to change can be overwhelming.
For those who have worked through the challenges and overcome the initial hurdles, however, the results can be transformational. With the right data analytics tools geared to the way scientists and engineers think about their data, tasks that previously took days, weeks or even months can be reduced to minutes and hours. The benefits include greater predictability, improved efficiency, and the freedom to focus on the next challenge. It is enough to make proponents of data analytics passionate about the possibilities.
Looking at the Data
One example is an experience I had while working on a root cause investigation for an issue with yield in a vaccine manufacturing process. The problem had been ongoing for some time, and analytics was almost the last resort for finding a solution. The various teams involved all had theories: The downstream scientists were convinced the problem was in the chromatography steps; the upstream scientists thought a raw material change was the culprit; the lab people didn’t think it was their fault because they had made some improvements to analytical testing at around the same time. The operators running the process had noticed some changes, but no one was listening to them. The organization had reached an impasse.
The decision was made to gather the data and put it into electronic format, then begin to apply analytics to uncover the best way forward. By looking at the data and acting on what it revealed, rather than searching for evidence to support existing theories, the organization achieved a break through. The early results of the analysis changed everything and the whole investigation turned at that moment. The data helped all to focus. Analytics allowed the teams to build consensus and to start listening to the process technicians and what they had been trying to tell them all along.
In this kind of situation, it is often those who shout loudest who get heard. Data analytics can help to level the playing field and allow members of the team to relinquish erroneous beliefs that may be holding the whole operation back. By trusting the data and allowing analytics to be a “myth killer,” greater advances come within reach.
Solving a high-profile problem for your organization will certainly raise the profile of an analytics approach. People should take risks and tackle the big problems with data analytics, rather than the projects that allow them to hide. High-profile projects are more likely to get resourcing and management attention, which in turn help to drive momentum and organizational change when people start to see the results.
In the vaccine manufacturing example mentioned earlier, the problem with yield had been ongoing for 2 years and was solved in 3 weeks once the data was available. There are many other examples of massive reductions in total development time — the total time it takes to reach a stated goal — achieved through data analytics. The scale of change may be dramatic. The resulting freedom to innovate, to move on to solving other problems, to adopt proactive approaches rather than reactive, can make all the difference to the motivation levels of the teams involved. Data analytics can cut the time required for research and development, helping R&D to support twice as many products, and bring them to market twice as fast. And because knowledge accumulates, researchers can innovate more predictably over time.
The clarity and certainty of data-driven decision making can transform the outlook of a team and help to drive a problem-solving culture based on analytics excellence. There may be an initial investment of time required to plan the approach, but if data is available in electronic format it can be used and analyzed effectively. It is only by starting to ask questions of your data that you can understand whether you have the right data in the first place.
It is also advisable not to use lack of data availability as an excuse. The results can quickly convince colleagues, teams and the wider organization that there are significant time savings and cost reductions to be achieved if they embrace new techniques. Sometimes it can be the production of a single compelling graph, or an interactive visualization of the problem, that helps generate consensus. Data analytics can help to focus the team and deliver the certainty and predictability to reassure the wider organization. It can also reduce the stress levels of the individuals involved, leading to a more productive, fulfilled team able to reach its potential.
Delivering results and solving high-profile problems can rapidly generate interest from the wider organization. Management and leadership teams will want to understand how to capture the value of a data analytics approach once they start to see its positive impact. Scaling data analytics to help transform an organization requires the right tools. Selecting a tool that guides non-statisticians through appropriate ways of analyzing their data helps to ensure the willing adoption of new ways of working. In a head-to-head comparison, a tool designed around the way scientists and engineers think about data in the first place is a winning choice. Avoiding a silo mentality is also crucial. By enabling researchers and developers to conduct their own analysis, organizations can scale from the ground up. By putting data analytics capabilities into the hands of the subject matter experts — those who understand the data, its origins and its quality — organizations can scale capabilities more rapidly to meet the growing demand for analytics excellence.