Let’s Play Moneyball: 5 Industries That Should Bet on Data Analytics


“Moneyball,” written by Michael Lewis and published in 2003, recounted how the Oakland Athletics baseball team used statistics and sophisticated analytical techniques to decide which players to sign—something no other team was doing at the time. These decisions, which were previously made by baseball scouts who relied on experience and intuition, allowed a small team with limited resources to compete against organizations with far deeper pockets.

The opportunities for Moneyball equivalents in the business world are enormous. More and more organizations are collecting data in unprecedented quantities (often called “big data”). By analyzing it using algorithms and statistical techniques, smaller companies can reveal previously undiscovered patterns and insights to help level the playing field between themselves and larger organizations.

We interviewed several business intelligence experts to get their opinion on which industries are ripe to be “Moneyballed,” or have the most to gain from the application of data analysis.

Sales: Identify Employee Skills That Impact Revenue Most

Companies are constantly on the lookout for ways to improve employee sales skills, and analyzing data surrounding these skills can help managers understand which ones have the greatest effect on revenue.

Joe Brooks is COO and general manager of Zapoint, a company that uses data analysis to help organizations engage, retain and develop their workforces. He gives an example of how one of Zapoint’s clients was able to analyze employee sales skills to gain a better understanding of which are most important for training.

Employees completed an assessment where they were asked to rank how important a series of sales skills were for their current role. Below are three examples:

  • Identify measurable results: LOW
  • Focus on value creation: MEDIUM
  • Sales presentation: HIGH

Based on these results, the company structured their training program to place the greatest emphasis on the highest-rated skills. After employees completed the training program, their performance was compared against their performance prior to the program, with surprising results.

The assessment results suggested that employees whose performance increased in the skill areas they considered most important would have the greatest impact on revenue. In other words, it was expected that those who improved their sales presentation skills would impact revenue more than those who improved in identifying measurable results.

But the data told a different story. The company found the most important skills in relation to revenue were actually as follows:

  • Identify measurable results: MEDIUM
  • Focus on value creation: HIGH
  • Sales presentation: LOW

“In this case, the analysis showed that investing time and money on emphasizing value creation would likely have the biggest impact on sales revenue,” explains Brooks. After evaluating the data, the company revamped their training program to focus on the skills shown to have the greatest impact. It now conducts ongoing analyses to uncover similar patterns.

It’s clear that widely held beliefs concerning the most essential sales skills may not always hold true under the lens of data analysis. Smaller organizations that mine their “big” data have a chance to get a leg up on larger organizations by identifying the most important skills and adjusting their training accordingly to focus on them.

Pharmaceuticals: Discover New Uses for Drugs

Identifying drugs to treat a variety of diseases is another way Moneyball tactics are being used in the business world. Amar Shan is the director of product marketing at YarcData. One of his customers is a Seattle-based drug company that uses data related to the genetics of cancer and non-cancer cells to test drug effectiveness.

In an attempt to identify previously undiscovered uses for drugs, the company implemented a data analysis system that synthesizes cancer and non-cancer cell data with data from research published on Medline. Rather than having to conduct their own experiments, researchers use this combined database to test their hypotheses against all known data that exists.

Within six weeks of installing the system, the company discovered a gene mutation responsible for a particular type of breast cancer that is also directly related to HIV; a connection no one had made before. Hints of a possible connection did exist previously—an HIV patient with breast cancer had taken the drug as part of her HIV treatment plan and unexpectedly seen an improvement in her cancer—but no one had followed up on it.

Now that the data analysis system has revealed the true extent of this connection, the drug’s effectiveness against breast cancer is being tested, and the results look “very promising.” Shan says the company is now using the same analytical techniques with several dozen drug compounds in the hopes of unveiling similar breakthroughs.

This example highlights a key opportunity for smaller pharmaceutical companies looking to compete with larger, more established ones. By combining their own data with the vast amount of research that already exists and using data analysis to identify potential connections, organizations can uncover new uses for drugs that previously would have gone unnoticed.

Agriculture: Determine Optimal Facility Locations

Charles Linville, Ph.D., is president and founder of Ploughman Analytics. His work focuses on the agricultural industry, and he says analytics are useful at many points along the agricultural value chain. Beneficiaries include both the farmers who harvest crops and the companies that process these crops into food products alike.

Linville helps companies such as grain merchandisers identify the best locations to build new facilities. In the past, this decision was complicated by the enormity of options available. Today, advanced analytics helps distill these options so companies can make more informed choices.

Linville uses analytics to combine several types of data, including:

  • Where crops are processed;
  • How much of each crop is processed;
  • Where processing or transportation facilities are located;
  • The price of each crop offered at each of these facilities; and
  • The cost of transporting the crop.

By evaluating these factors together, he helps customers evaluate the sourcing and distribution regions that will naturally emerge when different price scenarios are applied to different potential locations. In one model, for example, a mere $0.05 change in offered price resulted in a whopping 50 percent shift in expected volume at a grain merchandising facility.

Thanks to significant advances in computer technology, this data can also be processed more quickly than ever. Ten years ago, the analysis Linville conducts had to be done by hand and took weeks. Now, it can be completed in a fraction of the time, which helps companies make critical decisions more quickly and stay a step ahead of their competition.

Agricultural companies can benefit from using data analysis techniques such as these to identify optimal facility locations so money and other resources aren’t wasted on projects that will deliver sub-par results. This is especially an advantage in today’s environment, where the continual depletion of resources requires an increasingly fast and more adaptable approach in order to remain profitable.

Auto Insurance: Use New Variables to Assess Risk

Statistical decision making is nothing new to the insurance industry, where actuaries have been using numbers and analytics to predict the financial cost of risk and uncertainty for a long time. But, according to Bruce Winterburn, VP of industry relations at Vertafore, an insurance software provider, the techniques have gotten increasingly sophisticated in recent years.

In the past, variables used to determine auto insurance rates were fairly limited—sex, age, automobile, driving record. Since then, it has been found that other variables seemingly unrelated to driving, such as occupation, financial stability and how often one moves to a new address, are also good indicators of risk.

When Winterburn was in his mid-40s, for example, he wanted to purchase a Corvette and received quotes from multiple insurance carriers. Not surprisingly, given that he was buying a sports car, most were fairly high. But one company quoted him a rate that was a quarter of the others.

Winterburn believes the company looked at a large variety of variables, such as education and occupation, and (correctly) decided his risk was relatively low. As a result, they were rewarded with his business. Even though he eventually sold the car, Winterburn continues to purchase insurance from the same company.

He says that Moneyball techniques are changing some fundamentals within the insurance industry. “The way insurance has worked in the past is, you try to calculate probable risk and spread that risk across an entire group,” Winterburn explains. “There is an assumption that those with higher risk pay less and those with lower risk pay more.”

But with big data and better analytical techniques, insurance companies can make risk pools smaller and smaller, possibly even narrowing them down to a single person. “It will be interesting to see how far regulatory bodies will let this go,” Winterburn says. “Those with higher risk may be priced out of the market.”

The application of these analytics is especially promising for smaller insurance companies who currently may not be able to compete with ones like Geico, Allstate or State Farm. The ability to pinpoint the exact factors that determine risk is a key advantage, as it allows them to correctly identify and insure lower-risk drivers who are less of a liability to the company and thus more profitable to retain.

Nonprofits: Uncover More Accurate Donor Giving Trends

It is not just the for-profit business world taking advantage of Moneyball techniques; non-profits have also entered the game. Richard Becker, president of Target Analytics, a division of Blackbaud, says organizations are using advanced analytics to boost fundraising efforts. By analyzing donor data, nonprofits can better understand how to acquire, renew, convert and upgrade prospects.

Becker says that one of the greatest benefits of analytics is the ability to more adequately stratify or segment donor prospects based on their value or suggested target ask amount. For example, fundraisers have historically relied too heavily on segmentation based on identifiable wealth and assets. “This approach is often limiting as (a) many wealthy individuals conceal their wealth and assets and (b) wealth is an indicator of capacity to give, not likelihood to give,” he explains.

Now, however, nonprofits are using analytics to identify individuals with both the capacity and likelihood to give. Becker says that analytics also allows organizations to use different measurements of success than before.

In the past, for example, the success of direct marketing campaigns was based on response rate. As a result, campaigns with many responses from low-dollar donors with low retention rates were deemed successful, even if they didn’t raise a significant amount of money. Analytics can help change this by identifying those donors likely to give larger amounts and who will continue to contribute over time, which will increase fundraising rates.

Smaller, less-established non-profit organizations have much to gain from these analytical techniques, as they often must compete against much larger organizations that attract attention—and donations—from numerous individuals and corporations alike. The ability to correctly target the most generous donors the first time around can help increase campaign success rates and ensure these organizations not only stay afloat, but grow.

Moneyball techniques may not consistently lead to success in the business world (the Oakland A’s didn’t win the World Series, even with their use of analytics). And as more and more companies deploy sophisticated analytics, the competitive advantage of this strategy may wane. But right now, employing a Moneyball approach can give companies, especially smaller ones, the advanced insight needed to lift them above their competition.

As Brooks points out, “With the interest in and proliferation of big data solutions, more applications are now becoming available that allow companies to apply a tried and tested model off the shelf, so there is no reason why [data analysis] should not be just as applicable for smaller companies.”

Thumbnail image created by txnicole.

Share this post:  
Alan S. Horowitz

About the Author

Alan S. Horowitz is a contributor to Software Advice.

Connect with Alan S. Horowitz via: