Best practices, Building data pipelines

How to improve your data analytics factory

Would we approach the practice of data analytics differently if we valued our data appropriately?

In his recent article, “Your Company’s Data May Be Worth More Than Your Company,” analytics author and educator Doug Laney reports that during COVID, when both United and American Airlines secured multi-billion dollar loans by collateralizing their customer loyalty programs, their customer data was appraised at two to three times more than the market value of the companies themselves. 

How is that possible? Whereas lenders (and investors) increasingly recognize the value of data, accounting rules still do not generally permit businesses from recognizing and reporting its value on their balance sheets. But that should not stop businesses from recognizing it’s true value.

In an age where data is the new oil–the world’s most valuable resource–this  accounting mismatch can have severe consequences. Writes Laney: “A failure to quantify the actual and potential value of your company’s data can lead to complacency in the way it is governed, integrated, transformed and made accessible. And this in turn results in a diminished ability to generate economic benefits from the data.” 

The Googles and Amazons of the world notwithstanding, this is exactly what we see in many data analytics programs. If data is the new oil, then analytics is the factory by which we create value from it. But in many organizations, the factory is not a very productive one. We see large, ongoing infrastructure investments; high IT cost burden, and long production times to get to simple, and often unsatisfactory outputs. It’s run less like a factory and more like a skunk works– a project developed by a relatively small group of people (in analytics, usually highly trained specialists) primarily for the sake of radical innovation.

But data analytics today is not radical innovation. It’s necessary for survival.

The most valuable companies today derive much of their value not from factories or planes but from intangible assets–intellectual property, software, copyrights, and databases. This is what the future should look like for most businesses. As Michael Mauboussion and Dan Callahan of Consilient Research at Morgan Stanley point out in their brief, “One Job: Expectations and the role of intangible investments,” investment in intangible assets now far outstrips investment in tangible assets. Back in 1979, tangible investment was 1.7 times that of intangible investment. By 2017 intangible investment was 1.4 times that of tangible investment. This is a major shift.

The ability to make use of your data to improve the customer experience, and generate new products and intellectual property, and ultimately increase the value of the entire company is not only fundamental to your business, but if you’re not able to do that, you’re actually taking value away from your company. 

Mauboussin and Callahan cite the business school example of an imaginary company called Top Toaster. Top Toaster’s return on investment drops as competitors come along and drive incremental returns toward the cost of capital. Once their return on incremental invested capital is equal to the cost of capital, their investment is not generating new value. “This is in the future of almost all companies,” write Mauboussin and Callahan. “Sometimes this reality is near and sometimes it’s distant.”

To postpone the fate of Top Toaster, they write, a company’s investments must earn a return higher than the opportunity cost of capital, and today that means investing in intangible assets, for two main reasons: scalability and exclusivity.

Intangible assets are what economists call “non-rival” goods because they can be used by more than one person at a time. They also tend to be “excludable,” i.e. the owner can prevent others from using them through mechanisms such as patents and copyrights.

Companies that are entirely data-driven are generating new IP and improving on brand through effective data and analytics practices.

So what can companies do to improve the efficiency of their data analytics factory?

  • To start with, find ways to do more with less. You have to have systems that are flexible and offer as much of a total solution as possible in order to shorten the production line to speed up output.
  • Allow experts throughout the business–not just IT professionals and data scientists and analysts–to efficiently and effectively get access to data that lets them leverage their knowledge to ask questions, build new analytics, make new discoveries, and satisfy and nurture their curiosity.
  • Future proof your analytics practice so that when new data ecosystems start to emerge, you can leverage them in a way that’s meaningful, without putting undue cost burden on the organization to address that data ecosystem change. 

Why haven’t we been able to get there? 

The tools that we have tied ourselves to haven’t focused enough on the problem of end user access. The promise of self-service analytics and data access has been rampant in the industry for many, many years. But it’s been more like the Simpsons’ monorail–the “revolutionary vehicle of the future.” That’s what people have been investing in but not getting. 

We keep tweaking the infrastructure, but it is time to attack the problem in a different way. 

Before we invest in building the giant data factory, let’s take the interim step of connecting the data and letting people explore it (with appropriate controls) to get a sense of the forward value.

New data sources are always emerging, but not all of it is equally valuable. A more diverse group of people, with greater agility to work with the data can incubate insight and help you instantiate the blueprint for the bigger factory.

It does require continuous investment, but the investment shifts toward enabling exploration and engendering a more data-curious organization. This is the way to drive down the total cost of ownership for analytics infrastructure while increasing ROI.

When you’re looking at the classic understanding of a company and tallying up a firm’s assets such as buildings, land, planes, factories, etc., all of those things can be appropriately priced and accounted for. It’s relatively easy to predict how incremental investment in more of the same will increase output and revenue. 

It’s not so straightforward with investments in intangible assets such as data and analytics, and accounting rules have not kept up with the dramatic growth of intangible assets. Companies can’t afford to wait for that to happen. They can’t wait for a crisis to begin monetizing their data. And they can’t continue to invest in underproducing analytics factories. They must put data in the hands of everyone in the company to explore, and make generating insights an all hands on deck exercise.


Prove out the value of Incorta today. Give us your toughest data challenges and see immediate results.