How to plan great explanations of a data science model for business management

Explaining data models and all the assumptions can be complicated, but there are ways to simplify details. Here are 5 tips on how to make simple explanations great.

No matter where you are in the data science journey, you will often hear about data but need to be made aware of how to best explain the ideas from it. This is especially tricky when working with managers who need to be made aware of the nuances of the data. Well, the best way to get the message across to those managers is to plan the explanation. Doing so highlights how to best illustrate the results of a model.

There’s a saying that the answers are often never as important as the right question. So, the step in planning a good explanation focuses on the right question that will drive discussions, insightful answers, and meaningful tasks.

Start with the audience needs

Identify what to prioritize and how it will be delivered. An audience is the most obvious starting point, but the purpose of a data model is to address or describe in answer to a question being posed. So, you need to think about how to best frame audience needs to help you plan your model and plan how your model relates to answer the key questions. Many times, I use presentations as a way of distilling information into coherent topics. Thus, you may want to think about having bullet points.

Prioritize The Right Data

The second step is to note how the data is prioritized in the model itself. Doing so critiques what the model is supposed to do and what data sources are involved. Data is meant to represent real-world experiences, so you need to explain how certain data is important to the model. For example, many times people have explained housing data sets for their model. That means you can imagine a table with columns of data rows each describing a house and the columns with each house attributes like square footage, number of bathrooms, number of bedrooms, scoring of neighborhoods, and so forth. Any detail explaining a variable (column) to real-world descriptions is helpful. It will also make…