Chapter 2. AdventureWorks Regression

Imagine you're a business developer at AdventureWorks, a bicycle manufacturing company based in Seattle, Washington. You are responsible for three applications that run at the top of a single SQL Server instance. The applications are:

• A customer ordering website with a section for direct customer sales and another section for resellers to buy in bulk
• A desktop inventory control management application
• A reporting solution using Power BI as a frontend

All three of these applications share similar characteristics:

• They are database-first applications where their primary role is to wireframe the database
• They are all .NET applications that use standard Microsoft templating and frameworks, such as MVC for the website and Entity Frameworks for both web and desktop solutions

One day, your boss calls you into her office and says, "we are concerned about the reseller's section of the website. We've noticed through some basic charting in the Power BI that many resellers are dropping their order depending on the average customer reviews of the product.

Here is the one of the charts we are looking at:

Obviously, if we can prevent people from doing this, we will maximize sales. We want to maximize our existing code assets, so your solution needs to integrate with the existing website and we want our customers to experience the same look and feel they currently have."

This is what the current webpage looks like:

You tell your boss that you will take a look, think about it for a couple of days, and come up with some ideas. Inside, you are thrilled because this will take you out of the traditional role of web dev and into data science. After researching some different machine learning techniques, you settle on using a simple regression to help achieve this goal.