Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

Beyond algorithms: The philosophy of data ownership

Understand what insights your data could generate, or it remains a potential rather than an asset.

[Photo: Getty Images]

Profile photo of Simin Cai

What if the barrier to extracting value from AI isn’t computing power or algorithms? Instead, that barrier is not knowing what assets you actually possess.

As someone who approaches business problems through a philosophical lens, I see this question in terms of ontology, epistemology, and axiology: what you have, how you know it, and what it’s worth. A business might be sitting on internal insights that could save millions, or external licensing opportunities, and not even know it.

Think of data like inherited artwork. You might know you have a painting, but is it a reproduction or an original? What is its artistic and historical significance? What is it worth in today’s market?

Before your business can profit from its data, you need that systematic appraisal process. Let’s walk through these three phases because they will reveal opportunities hiding in plain sight.

ONTOLOGY: WHAT YOU HAVE

The simple principle behind identifying what data you actually have is that you cannot own what you cannot see. But we are not talking about taking inventory.

For example, a retail chain tracks customer purchases, website clicks, return patterns, seasonal fluctuations, and employee scheduling data. The data exists across multiple systems, but without clear understanding of what specific insights it could generate, it remains as potential rather than an asset.

This is the perception problem.

For years, UPS collected vehicle telemetry from delivery trucks, such as GPS coordinates, speed data, stop information, and package tracking details. Initially, the data served logistics management. But UPS’s operations research group recognized it could drive route optimization, leading to the development of its On-Road Integrated Optimization and Navigation (ORION).

The result? Savings from reduced miles and fuel are projected to hit $400 million in 2025.

To adopt an ontological frame, start by cataloging what questions your data could potentially answer. List the business problems each dataset might solve, both for your organization and for external partners who might license it.

Once you can identify what you have, the next question is how expert analysis can unlock its meaning.

EPISTEMOLOGY: HOW YOU KNOW IT

The British supermarket chain Tesco used to collect millions of transaction records daily in the early 1990s, but they only served inventory management and promotions. They were effectively just receipts. To make more of their data, Tesco decided to partner with academic researchers from a startup called Dunnhumby.

When Dunnhumby’s data scientists presented their insights about shopping behavior and customer segments to Tesco’s board, then chairman, Lord Ian MacLaurin declared: “What scares me is that you know more about my customers after three months than I know after 30 years.”

The choice not to build an internal analytics team proved its value. In fact, the loyalty program born out of the research and analytics, called Clubcard, was pioneering. Tesco soon overtook Sainsbury’s, and by year three, Clubcard had doubled the company’s grocery market share.

Today, AI has dramatically amplified these analytical capabilities. Machine learning algorithms can identify patterns that would be invisible to traditional statistical methods. But the principle remains the same—meaning is never absolute.

The Tesco example shows that the same dataset, viewed through a different methodology or discipline, may reveal entirely different insights, which may potentially bring significant commercial value. So, widen your lens. Partnering with universities or external researchers can bring methodologies from statistics, behavioral science, or operations research that reveal patterns your internal teams might miss.

AXIOLOGY: WHAT IT’S WORTH

Proper data valuation requires looking at both internal cost savings and external market opportunities. But sometimes, value can be found in both domains.

Like UPS, Rolls-Royce was already collecting telemetry as part of its normal operations in the 1990s. Sensor data from aircraft engines such as temperature readings, vibration patterns, and fuel consumption metrics all helped predict engine failures before they occurred.

But then the aerospace company parlayed predictive intelligence into TotalCare service agreements, which now drive growth in its aftermarket revenue. Instead of selling engines, Rolls-Royce now guarantees engine availability. Airlines pay based on flight hours while Rolls-Royce assumes all maintenance risk.

Farm equipment giant John Deere followed a similar path, turning agriculture data into subscription revenue. AI-powered software now analyzes soil conditions, crop yields, and planting patterns to generate predictive insights that farmers pay for monthly. Deere is predicting software revenues will represent 10% of total company income by 2030.

Value doesn’t always mean marketable products. Internal datasets on employee well-being may never be sold, but they can strengthen culture, resilience, and long-term sustainability. Sometimes the greatest value is cultural or ethical before it is financial.

BEYOND THE ALGORITHM

AI is often discussed as if algorithms are the only story. But the opportunity represented by unrealized datasets is in reshaping the way we see being, knowing, and valuing within our own organizations.

In art, brush marks can be seen as a flaw or the signature of a noted painter. Similarly, what appears to be operational noise in your data might actually be the fingerprint of untapped value, if you know how—and where—to look.

Simin Cai is president and CEO at Go!Foton.