Brief

With Advanced Analytics, It's People (Not Data) That Stand in the Way of Change

Companies that take a one-team approach to behavior change will see transformational results with advanced analytics.

By Chris Brahm, Lori Sherer, Richard Fleming and Briana Bennett

First published on agosto 25, 2017
min read


Only five years ago, a World Economic Forum report predicted that data would become its own kind of economic asset, able to hold value similar to gold or dollars. Ever since then, companies have been rushing to horde and harness their cut of the more than 23 zettabytes of data available—an amount that’s expected to nearly double by 2020, according to IDC. Among the 334 executives that Bain recently surveyed, more than two-thirds said that their companies were investing heavily in data and analytics. Not surprisingly, 40% expect to see “significantly positive” returns on their investments, with another 8% going as far as predicting “transformational” results. Their optimism isn’t unfounded—companies from UPS to USAA are mobilizing advanced analytics to great effect. Even so, 30% of these executives said that they lack a clear strategy for embedding data and analytics in their companies. And despite the best intentions of the 70% whose companies have strategies, many will lose their way with their data because of one simple reason: people. A company can have the most sophisticated tools and the most brilliant data scientists, but its efforts will fail without the behavioral changes necessary to support decision making and action. Consider the recent experience of a global food manufacturer. With the goal of reducing spoilage and shipping costs, the company invested in an algorithm that aimed to predict demand at thousands of distribution points. Ideally, the algorithm would match deliveries with expected sales to keep stale food from ending up in trash cans. However, the company never created incentives to motivate stores to order based on demand predictions. And worse, the algorithm’s forecasts turned out to be inaccurate, further hobbling the effort. Lori Sherer, a partner with Bain's Advanced Analytics practice, shares three strategies for companies looking to productively invest in data and analytics to drive business performance. In our experience, the journey to lasting results with advanced analytics starts with a desire to fix a critical business problem. From there, companies choose their data engineering and data science approaches before moving on to deployment and then adoption. However, a company’s advanced analytics efforts will inevitably fall short if a company doesn’t take the necessary steps to change behavior (see Figure 1). This one-team approach brings together stakeholders from all areas of a company to build early support for an advanced analytics initiative. Getting there involves four crucial steps: co-creating “the beach” by bringing employees from different parts of the company together to develop the company’s vision of how it will benefit from advanced analytics; engaging the sponsorship spine by enlisting critical leaders at every level of the company who can help motivate the right behaviors from employees; orchestrating for success so that leaders and employees can anticipate potential challenges; and measuring and inspiring adoption so that companies can change course if necessary. The human aspect of advanced analytics can’t be understated—the vast majority of business processes are still governed and carried out by people. In fact, behavior change issues account for the five most common reasons that we see in disappointing advanced analytics initiatives. Reason No. 1: The front line isn’t committed to using data analytics. Perhaps frontline employees weren’t engaged at the start of a company’s forays into advanced analytics, or they don’t see the value of the data. In either case, it’s likely that leaders haven’t communicated the beach—that is, the company’s vision for how frontline employees would get the most from advanced analytics. Swiss Life encountered such doubts when it introduced new analytics tools to help employees increase sales leads. The insurer overcame the naysayers by piloting the effort in an underperforming business unit that used the tool to beat its sales targets. The insurer publicized the victory across the company, which helped recruit support for the effort. Reason No. 2: The data science and business teams aren’t communicating. Too often, data science teams dump insights over the transom and let business units make sense of it. That approach rarely works. At leading companies, Agile and cross-functional teams tackle a specific problem, along with input from the employees closest to the issue. Data scientists might report to a central leader who guides companywide analytics efforts, but they immerse themselves in their designated business units, allowing them to stay close to the products and customers they serve. That approach has worked for data leaders Netflix and Airbnb. And when it’s time to share internal analytics developments more broadly, leading companies turn to their best communicators—often their marketing teams—who can convey opportunities with compelling visuals and easy-to-grasp language. Reason No. 3: The data solutions aren’t user-friendly. “Black box” or overly academic data solutions—those that rely on opaque, generic back-end technology—seem like an easy way for companies to catch up in the advanced analytics game. But these tools can provide clunky, overly complicated insights that are impossible for employees to deploy at scale. The best systems turn complex data into simple visuals and scores that enable quick action. Think about FICO, which takes a consumer’s complex loan history and calculates a simple measure of creditworthiness that banks have been using to make lending decisions for more than 25 years. The scores are so effective that banks have started to share them with customers, allowing them to take steps to improve their financial health. Reason No. 4: The data users are not prepared to change their behavior. Adopting new processes and tools can intimidate even the most seasoned employees. Naturally, companies deploying Big Data solutions need to provide constant training and coaching to not only teach employees how to use new technology but also to understand the decision-making implications at every level of the company. Companies that do it best establish a strong sponsorship spine that can help motivate the right behaviors from employees. Without a sponsorship spine to support deployment and adoption, an insurer’s senior underwriters might feel threatened by tools that calculate predictive scores for lending decisions rather than rely on their expertise. A telecom company that provides detailed customer feedback to call center employees might never reap the benefits of personalized service, a feature that’s quickly becoming a basic service expectation. Reason No. 5: A company fails to reinforce and monitor critical behavior changes. Supervisors can provide frontline employees with new tools and data, but real change only takes hold with clear incentives and strong feedback loops that allow users to flag problems to analytics teams early and often. Consider a bank that gives detailed analytics about customer experience trends and cross-selling opportunities to branch managers, who then use that data to improve one measure at the expense of the other. If the bank had made it more appealing for employees to choose new behaviors instead of old and had provided positive reinforcement throughout the process, it might have gained on both fronts. Behavior change is often the hardest part of improving a company’s performance on any dimension. It’s the reason why only 12% of change efforts achieve or exceed a company’s expectations and 38% fail by wide margin. Many of the companies making huge investments in advanced analytics will be disappointed to discover that data tools alone aren’t enough to grow a company’s fortunes. However, companies that take a one-team approach to behavior change—by enlisting sponsors, creating their ideal vision, orchestrating for success and measuring progress—set themselves up for transformational results. Chris Brahm is a Bain & Company partner who leads the firm’s Global Advanced Analytics practice. Lori Sherer is also a partner with the Advanced Analytics practice. Both are based in San Francisco. Richard Fleming is a partner with Bain & Company in the New York office, and he leads Bain’s Americas Results Delivery^® practice. Briana Bennett is a manager on the Advanced Analytics practice in San Francisco. Results Delivery^® is a registered trademark of Bain & Company, Inc.

Only five years ago, a World Economic Forum report predicted that data would become its own kind of economic asset, able to hold value similar to gold or dollars. Ever since then, companies have been rushing to horde and harness their cut of the more than 23 zettabytes of data available—an amount that’s expected to nearly double by 2020, according to IDC.

Among the 334 executives that Bain recently surveyed, more than two-thirds said that their companies were investing heavily in data and analytics. Not surprisingly, 40% expect to see “significantly positive” returns on their investments, with another 8% going as far as predicting “transformational” results. Their optimism isn’t unfounded—companies from UPS to USAA are mobilizing advanced analytics to great effect.

Even so, 30% of these executives said that they lack a clear strategy for embedding data and analytics in their companies. And despite the best intentions of the 70% whose companies have strategies, many will lose their way with their data because of one simple reason: people. A company can have the most sophisticated tools and the most brilliant data scientists, but its efforts will fail without the behavioral changes necessary to support decision making and action.

Consider the recent experience of a global food manufacturer. With the goal of reducing spoilage and shipping costs, the company invested in an algorithm that aimed to predict demand at thousands of distribution points. Ideally, the algorithm would match deliveries with expected sales to keep stale food from ending up in trash cans. However, the company never created incentives to motivate stores to order based on demand predictions. And worse, the algorithm’s forecasts turned out to be inaccurate, further hobbling the effort.

Lori Sherer, a partner with Bain's Advanced Analytics practice, shares three strategies for companies looking to productively invest in data and analytics to drive business performance.

In our experience, the journey to lasting results with advanced analytics starts with a desire to fix a critical business problem. From there, companies choose their data engineering and data science approaches before moving on to deployment and then adoption.

However, a company’s advanced analytics efforts will inevitably fall short if a company doesn’t take the necessary steps to change behavior (see Figure 1). This one-team approach brings together stakeholders from all areas of a company to build early support for an advanced analytics initiative. Getting there involves four crucial steps:

co-creating “the beach” by bringing employees from different parts of the company together to develop the company’s vision of how it will benefit from advanced analytics;
engaging the sponsorship spine by enlisting critical leaders at every level of the company who can help motivate the right behaviors from employees;
orchestrating for success so that leaders and employees can anticipate potential challenges; and
measuring and inspiring adoption so that companies can change course if necessary.

The human aspect of advanced analytics can’t be understated—the vast majority of business processes are still governed and carried out by people. In fact, behavior change issues account for the five most common reasons that we see in disappointing advanced analytics initiatives.

Reason No. 1: The front line isn’t committed to using data analytics. Perhaps frontline employees weren’t engaged at the start of a company’s forays into advanced analytics, or they don’t see the value of the data. In either case, it’s likely that leaders haven’t communicated the beach—that is, the company’s vision for how frontline employees would get the most from advanced analytics. Swiss Life encountered such doubts when it introduced new analytics tools to help employees increase sales leads. The insurer overcame the naysayers by piloting the effort in an underperforming business unit that used the tool to beat its sales targets. The insurer publicized the victory across the company, which helped recruit support for the effort.

Reason No. 2: The data science and business teams aren’t communicating. Too often, data science teams dump insights over the transom and let business units make sense of it. That approach rarely works. At leading companies, Agile and cross-functional teams tackle a specific problem, along with input from the employees closest to the issue. Data scientists might report to a central leader who guides companywide analytics efforts, but they immerse themselves in their designated business units, allowing them to stay close to the products and customers they serve. That approach has worked for data leaders Netflix and Airbnb. And when it’s time to share internal analytics developments more broadly, leading companies turn to their best communicators—often their marketing teams—who can convey opportunities with compelling visuals and easy-to-grasp language.

Reason No. 3: The data solutions aren’t user-friendly. “Black box” or overly academic data solutions—those that rely on opaque, generic back-end technology—seem like an easy way for companies to catch up in the advanced analytics game. But these tools can provide clunky, overly complicated insights that are impossible for employees to deploy at scale. The best systems turn complex data into simple visuals and scores that enable quick action. Think about FICO, which takes a consumer’s complex loan history and calculates a simple measure of creditworthiness that banks have been using to make lending decisions for more than 25 years. The scores are so effective that banks have started to share them with customers, allowing them to take steps to improve their financial health.

Reason No. 4: The data users are not prepared to change their behavior. Adopting new processes and tools can intimidate even the most seasoned employees. Naturally, companies deploying Big Data solutions need to provide constant training and coaching to not only teach employees how to use new technology but also to understand the decision-making implications at every level of the company. Companies that do it best establish a strong sponsorship spine that can help motivate the right behaviors from employees. Without a sponsorship spine to support deployment and adoption, an insurer’s senior underwriters might feel threatened by tools that calculate predictive scores for lending decisions rather than rely on their expertise. A telecom company that provides detailed customer feedback to call center employees might never reap the benefits of personalized service, a feature that’s quickly becoming a basic service expectation.

Reason No. 5: A company fails to reinforce and monitor critical behavior changes. Supervisors can provide frontline employees with new tools and data, but real change only takes hold with clear incentives and strong feedback loops that allow users to flag problems to analytics teams early and often. Consider a bank that gives detailed analytics about customer experience trends and cross-selling opportunities to branch managers, who then use that data to improve one measure at the expense of the other. If the bank had made it more appealing for employees to choose new behaviors instead of old and had provided positive reinforcement throughout the process, it might have gained on both fronts.

Behavior change is often the hardest part of improving a company’s performance on any dimension. It’s the reason why only 12% of change efforts achieve or exceed a company’s expectations and 38% fail by wide margin. Many of the companies making huge investments in advanced analytics will be disappointed to discover that data tools alone aren’t enough to grow a company’s fortunes. However, companies that take a one-team approach to behavior change—by enlisting sponsors, creating their ideal vision, orchestrating for success and measuring progress—set themselves up for transformational results.

Chris Brahm is a Bain & Company partner who leads the firm’s Global Advanced Analytics practice. Lori Sherer is also a partner with the Advanced Analytics practice. Both are based in San Francisco. Richard Fleming is a partner with Bain & Company in the New York office, and he leads Bain’s Americas Results Delivery^® practice. Briana Bennett is a manager on the Advanced Analytics practice in San Francisco.

Results Delivery^® is a registered trademark of Bain & Company, Inc.