Using Analytics to Make Better Decisions
by Blair White
Knowledge isn’t power – the application of knowledge is power. PCCA holds itself to a new standard by using industry data to help drive business decisions. Leveraging experience and intuition into something testable and repeatable is a continuous process. However, a recent re-evaluation of the cooperative’s risk management policy suggested that a powerful piece of the puzzle didn’t fit quite right. That’s when a reimagined analytics and data sciences initiative was born at PCCA.
“At times, we are prone to lean on our experience without validation,” said Kevin Brinkley, PCCA President and CEO. “Data can confirm our understanding of what occurred, or it can uncover evidence of previously unseen facts. For example, we believe our data assets have given us a better understanding of production outcomes based on multiple factors, not just weather. That knowledge helps us improve our navigation of markets.”
Why Data Sciences?
Using data sciences aims to provide grower-owners with the highest possible value for their cotton. Solid analysis helps the co-op make more informed decisions. Good decisions equal added value. Cotton is one of the most complex commodities in the world, which makes the industry that surrounds it an overflowing wellspring of data. Having all that data is good, but data alone is not always valuable.
“Having data is not the same thing as having information and having information is not the same as knowing what to do with it,” explained Chris Kramedjian, PCCA’s Director of Risk Management. “That’s the main point of what we are doing with data sciences and analytics. We are taking these large sums of information and figuring out what is signal and what is noise, what actually matters. But I think it is important to note that this isn’t a hard break from PCCA’s past. In fact, our past is our starting point. We begin at the intuitions and practices of our deeply experienced team. Their insights in how things work in terms of production, markets, qualities, etc., and try to develop it into a formalized understanding that we can quantify and perhaps even improve. A lot of things that have been very intuitive and experience-based before are still in use, but benefitting from rigorous analysis.”
How it Works
PCCA collects useable data from USDA, risk management agencies, cotton exchanges, trading platforms, weather outlets, and aggregated business data.
“Any kind of data that can show a real relationship to one of the variables that really matter to us is something we are going to try to take into account,” Kramedjian said. “Typically, we come to considering a data source based on the problems, questions and observations we’re trying to examine. From there we figure out the theory and how to test it, and then when we look for the data that can help us answer the question. If the data can answer the question or solve the problem, that’s great and we turn to putting the resulting analysis in front of the right people in the right way. Although it may not seem like it, discovering that certain data cannot answer a question as we thought it could is also helpful. Discovering that some piece of closely watched data is just noise means we can redirect our attention and our resources to things that do matter. ”
Weaving PCCA’s staff’s expertise into solid data analysis is not a one-person job and is not just based in the Sales Department. It is a company-wide effort that requires skills from multiple departments, including Information Systems and Member Services. The applications of this practice also extend beyond marketing cotton.
“Intuitively, marketing is a priority focus area for data sciences, but it works in all aspects of our cooperative,” Brinkley said. “For example, the improvements we have made in warehousing using this technology have allowed us to make more efficient use of our space while enhancing our performance to customers. These are win-win scenarios for PCCA’s customers and grower-owners.”
Analyzing and drawing meaning from this data begins as a human-based effort. However, the investment in quantifying analysis has the added benefit of allowing automation where significant efficiencies are realized.
“Improvements in our understanding through machine learning and artificial intelligence means it takes less time to understand our data,” Brinkley explained. “Once we are satisfied with an analytical approach, we can use automation to produce results in real-time rather than a manual process that requires human intervention.”
Future Implications
Technological advancements are becoming more necessary as PCCA looks toward the future. As such, starting this initiative is no small feat for PCCA. Driving better business decisions begins with tenured employees but is completed by backing that strength with the power of numbers.
“Every single thing that we want to do employing data science analytics is serving the purpose of getting the growers the highest return from their cotton and their company,” Kramedjian said. “This initiative is directly aimed at serving PCCA’s mission. While a single farmer probably doesn’t have access to a team of people that have years of experience and decades of intuition piled up and the various skills it takes to add value through data sciences and analytics, he does have access to this cooperative. Our members can be confident that we’re actively expanding our capabilities wherever it can help us earn them better returns.”
