Wayne Winston has consulted for several Fortune 500 clients, developed the statistics tracking and rating system used by the Dallas Mavericks, is a two-time Jeopardy champion, has a PhD in Operations Research from Yale University, and holds numerous teaching awards. He's also written several books, including Mathletics, Microsoft Excel 2010 Data Analysis and Business Modeling, Practical Management Science, and Spreadsheet Modeling.
One of his books, Marketing Analytics, was included in the Top 25 Analytics Books to Read in 2016, and centers on data-driven techniques with Microsoft Excel. I had the chance to sit down with Winston and dig deeper into analytics while also getting some great business tips for young data scientists to get in the game. Check it out:
Tech.Co: Your book Marketing Analytics, focuses on Microsoft Excel to solve real world business problems. Why Microsoft Excel? Also, do you think the classic spreadsheet still has as much to offer data analysts and business owners as the newer marketing tools today, say for example, Google Analytics?
Wayne Winston: Most MBAs and business school grads know Excel so it is easy to teach them advanced Excel ideas that let them perform the key tasks in analytics. The learning curve is small. Also, in Excel you can see how the data looks and how transformations of the data look. This makes it easier, I think, to understand what the computer is doing.
I do think the “black belt” business analyst needs to be able to program in Python, R, and Java as well as have a deep knowledge of data mining. The problem is most business analysts do not have a programming background although they can pick it up through Lynda.com or other sites. Excel makes it easier for these business school grads to get started.
Excel add ins like XLMINER, XLSTAT and Stattools also make it easy to do most of what SAS, R, and other packages do. In summary I think the Excel approach takes more people further and more quickly than the other approaches to analytics. Most organizations, however, need analysts with strong Excel skills and the more advanced analytic skills.
Tech.Co: You, yourself have a PhD in Operations Research from Yale University, yet you write for those without such an extensive background in the field. Where do you believe business owners and data analysts need to start in order to better understand their customers and improve results?
Winston: I am biased but I think books like my Practical Management Science, Marketing Analytics, and Microsoft Excel Data Analysis and Business Modeling can get a person started. Then I think to become a black belt you need to be strong in programming, mathematical statistics, optimization, and simulation. There are many online masters programs that develop these competencies.
Tech.Co: You've consulted for many big-time organisations, including the Dallas Mavericks, Cisco, Microsoft, U.S. Army and Medtronics, but what can small to medium sized businesses learn from analytics, and do you believe on a smaller scale, analytics are just as important for small-time and newly established companies?
Winston: Yes. For example I go to my local bakery and they always run out of my favorite (rice pudding). If they were able to better forecast demand they would run out of less food and have less leftover. I helped our local bank forecast hourly customer counts and use queuing theory to determine the number of tellers needed each hour and schedule the tellers. I helped the local hospital pharmacy develop an Excel program to optimize inventory management for expensive drugs. Excel empowers small organizations with sufficient knowledge to develop tools that will enhance their profitability.
Tech.Co: You're perhaps best known for revolutionizing the way basketball teams rate performance but you've worked on analytics across multiple playing fields (pardon the pun), as an analytics expert do you believe data science is the same across the board or are you essentially doing something entirely different when you work for the New York Knicks vs. the U.S. Navy?
Winston: Most analysts are just involved with reporting but would like to branch out into more advanced analysis and decision-making. At most organizations my job is to improve analyst abilities to improve reporting capabilities. Then we work on improving the more analytic decision-making capabilities of the workforce.
At the Navy, I teach analysts how to use Excel to be more productive at work (mostly developing their reporting capabilities.) For example, one analyst needed to extract cell D1 from every 4th worksheet of a 200 worksheet workbook. She was doing this manually. By pointing to the cells The indirect function automates this process so it takes seconds. For the Knicks and Mavs we developed a new analytic method to rate players and lineups. We also developed a method to optimize lineups. Here we developed a decision making tool that provided new useful information and insights.
Tech.Co: What advice can you give to data analysts just starting out and do you give this same advice to all your students?
Winston: I would urge them to begin by getting good at Excel. Then, I would try and learn more statistics, optimization, and simulation. Then I would try and learn Java, R and Python and explore data mining. Again you probably need a masters in Data Science to develop the latter skills, but coursera.org offers an online certification in Data Science that might be a good way to go for many of your readers. Coursera also offers several Excel based analytics in business courses taught by Wharton faculty.