John H. Johnson, President and CEO at Edgeworth Economics, keynote speaker, and co-author of Everydata: The Misinformation Hidden in the Little Data You Consume Everyday, joins me on this episode of #Accelerate!
[2:42] John is a PhD economist with particular expertise in econometrics. Edgeworth Economics is data-driven and works by processing and explaining very large data sets. One large sector they serve is corporate litigation. John gives some detail.
[3:39] Much of John’s time is spent teaching these issues in courtrooms. His book is designed to bring this knowledge about real-world events to a larger audience, so people can make better decisions with data.
[4:25] The starting point is recognition. 90% of the world’s data was created in the last two years. People fear math. These two factors combine into the perfect storm for people to be misled and to misunderstand data.
[9:12] John suggests you should ask intelligent questions. To understand statistics, think about what went into producing the number.
[13:27] Even disciplined statisticians are prone to correlation confirmation bias. Consider, what questions you are trying to answer. Does the data give you enough complete information to answer the questions? What can it tell you?
[16:38] Large volumes of data may tell you something meaningful about your business and sales drivers. The application of this data doesn’t replace the interpersonal skills that are needed to connect and engage with clients.
[18:38] Making decisions on inapplicable correlations will not lead to the results you were expecting. Make sure you understand if the correlation is part of the causation.
[20:21] John comments on common sales stats, such as the Pareto distribution of sales to salespeople. Look behind the patterns. What could be causing them?
[23:10] Forecasting is only as good as the inputs and our ability to use past performance to predict the future. Hone in on the assumptions that underly the forecasting model. Forecasting is always probabilistic.
[28:45] Aggregate statistics about sales may be true, but drawing specifics from generalities is not trustworthy for any specific product and industry.
[30:34] John says managers should frame the question they want to answer and look for data that belongs to the question. Be aware where the data originates, and of assumptions under any analysis of it. Look at how it may, or may not apply.
[32:55] John emphasizes that data is a tool. It is a complement to decision-making. Use all the tools at your disposal. There is no substitute for thinking hard about these types of problems.