What’s the Value of my Data?  Today’s Most Critical Yet Hard to Answer Question

Today’s most critical question to which every organization should know the answer still goes unanswered in an age where the world’s most valuable resource is data.

I’m going to detail my approach to how I help my clients determine the value of their data. Here are some important theorems that underpin my approach:

I laid out this data valuation approach in “Building Value-driven Data Strategy: Use Case Approach – Part 2,” but I want to clarify further with an example in this blog.  So please do your homework and review that blog before diving into this blog.  Thanks!

In this blog, I will create a scenario for a high-tech manufacturer dependent upon a network of suppliers and contract manufacturers to build many of the components and subassemblies that comprise their sophisticated technology product.  This manufacturer has annual revenues in the $40B to $75B range and sells Business-to-Business.

In this scenario, this high-tech manufacturer already hired me (always a questionable move) to lead their organization through education, envisioning, and prioritization process (3-to-4-month engagement) to identify, validate, value, and prioritize their business and operational use cases around the following strategic business initiative:

Reduce inventory costs and components out-of-stock while improving supply chain and logistics predictability and quality.

Out of the envisioning workshop, we prioritized these three use cases:

Let’s dive into the data and analytics value attribution process for each use case.  This might get a bit tedious, but this probably be the best way to explain how the process works.

Use Case #1 seeks to Improve Vendor Quality by 5%, yielding $60M in annual savings by reducing customer product returns, reducing field service calls, reducing product reworks in the field, and improving customer referrals, Net Promote Scores (NPS), and Likelihood to Recommend (LTR).

The data science team, using the “Thinking Like a Data Scientist” process, determines that Use Case #1 requires 3 data sets (Customer Orders, Vendor Shipments, and Customer Product Returns) and two analytic modules (Anomaly Detection and Vendor Quality Score). See Figure 1.

Note: Determining the $60M in potential savings is an exercise that any financial analyst could complete.  No need to try to turn data scientists into financial analysts.

I’ve used a straight-line data valuation attribution model so that if the use case is worth $60M. There are 3 data sets and two analytic modules necessary to optimize that use case, then each data set gets one-third of the attributable value (1 ÷ 3 × $60M = $20M). Each of the two analytic modules gets one-half of the attributable value (1 ÷ 2 × $60M = $30M).  See Figure 2.