Editor’s note: This article was originally published in Harvard Business Review.
A recent survey found that businesses globally are spending close to $40 billion annually on technology and services for data analytics, increasing by 12% each year. And yet, in a survey of 64 C-level executives at large corporations, 72% said they had yet to forge a data culture, and about half admitted they were not competing effectively on data and analytics. What explains this disconnect?
Being a data-driven organization takes more than great technology and quality data. Like other aspects of digital transformation, it requires the right internal processes and culture — where the business properly guides incentives and takes steps to ensure that data is driving decisions appropriately. Failing to do this can lead to data misuse, which can be costly and hard to identify.
These organizations are “masquerading” as data-driven, meaning they have the data, technologies, and even the expertise, but their culture and processes are not aligned with those elements to produce the best outcomes. For example, data might be a part of every decision made, but employees may be making decisions first, then looking for data to back them up.
Factors like these explain the disconnect between investment levels and the disappointing results some companies report seeing. Businesses have more data than ever, but a culture rooted in top-down decision making and traditional tools like weekly reports and preconfigured dashboards means they cannot take full advantage of it.
Here are some common symptoms that your organization is masquerading as data driven, along with suggestions how to address them.
Your Employees are Making Decisions Based on the Tyranny of Averages
Treating customers, suppliers, and other stakeholders as a whole and making decisions based on averages can have harmful consequences. For example, my company works with a large auto manufacturer that has more than 500 suppliers globally. This company determined that the optimal time on aggregate to pay invoices was 30 days after they were submitted, and implemented this as standard.
In fact, account managers knew some suppliers didn’t require payment for 45 days, while others imposed a penalty after 25 days. As a result, some suppliers were unhappy and the business was losing money by paying some invoices earlier than necessary, and paying unnecessary late fees for others. The manufacturer addressed the issue, but it took months for it to be escalated up the chain and for the change to be made.
Your organization may be acting on data, but if the aggregate belies the reality underneath, the data is not giving the full picture. Nuances are lost, which can lead to lost income and damaged relationships.
Everyone Has Their Own Version of the Truth
When employees argue that “my truth is better than your truth,” it’s a sign you’re masquerading as data-driven. Each team may be acting on data, but if they have different information, they are bound to disagree and some may even be misled. The cause may be siloed data, where each team looks at their own slice of reality. Or there may be a lack of agreement about which data should drive a particular decision. Are we determining customer preferences by looking at service calls in a CRM system, or social data from a marketing tool? Getting stakeholders to agree on which data is important establishes a common source of truth to guide decisions and strategy.
More broadly, data should be available uniformly throughout an organization so all teams have access to the same information. The goal is outcomes, not ownership, and this may require a cultural shift that loosens the grip on data among senior managers.
Decisions Precede Data
Making decisions and then finding data to support them is a very human failing; we all tend to trust our own instincts. How many times have you made a decision and then looked for numbers to justify it to the board? There are two issues here: one concerns the sacrosanct way we’ve come to view data; the other how we communicate decisions to others.
First, recognize that instinct and experience are still critical in business. Some events — like the current COVID-19 crisis — are unprecedented, and historical charts are of no use in deciding how to act. The key is not to mislead others by applying numbers to justify a position. This sets a bad precedent when leaders should be setting an example.
If data is shared widely, you can lean on your teams — those with expertise in key areas — to uncover insights and opportunities. The leader’s job is to apply their experience to these insights and decide the right course. Ideas will not always start with the leader, but they will start with data. This may require suppressing the ego and accepting that in a data-driven world, ideas can originate anywhere, but leaders always have a critical role in deciding which course of action to take.
Employees Have Misguided Incentives
Data is often tied to bonuses and other rewards, but incentives backfire if they’re not applied judiciously. If sales leads know they’ll be rewarded for securing a second call, they’ll take the call even if it’s unlikely to generate a sale. People game the system; it’s human nature.
For an extreme case, look at the Wells Fargo employees who opened 2 million bank and credit card accounts without customers’ permission. This was an ethical lapse, but it was also driven by unrealistic goals — with repercussions for those who missed them.
Using data to drive incentives is appropriate, but targets should be used to motivate, not punish, or they will encourage bad behaviors. A customer service worker who averages 15 minutes to resolve a call instead of 12 may be considered “less productive,” but may also generate better results, greater customer satisfaction, and higher NPS scores, contributing more to the business overall. Focusing on the wrong data incentives will hurt a business.
Data holds tremendous potential to improve customer service, innovation, and efficiency, but organizations need the right environment to leverage its potential. Data should be distributed widely and uniformly, so that employees have a shared source of truth. And frontline workers must be empowered to act on data, with clear channels for sharing ideas and putting them into action quickly. This may require a significant cultural shift, but the opportunity is substantial — to capitalize on what’s possible and be a truly data-driven organization.