As a professor, I work with both sides of the data conversation. I teach MBA students to use data in business and data scientists to focus on impact, not accuracy. Outside the classroom, I advise and run workshops for both business and data leadership teams, trying to bring those two perspectives together.

No matter who I’m with, the same question keeps coming up: How do we use data to generate business value?

The problem is that business and data teams have spent the last decade answering that question from different viewpoints. And while both sides have made progress, true partnership has been harder to come by.

An uneven evolution

In the early days of the data boom, we placed most of the burden on business leaders. “Don’t get distracted by shiny tools,” we warned. “Define the problem first.” We told them that dashboards weren’t insights, models weren’t strategies, and more data didn’t mean better thinking. If they wanted value from data, they had to start by asking better questions.

It was good advice.

But as the data boom matured, organizations realized that clear business questions weren’t enough. If we wanted analytics to drive value, we had to build stronger technical teams, including data scientists and machine learning engineers. And we realized something else: we had spent years telling business leaders they needed a working knowledge of data science. Now we had to tell data scientists they needed a working knowledge of the business.

This shift in emphasis was necessary, but it didn’t go perfectly. We had told the data teams to make their work useful, usable, and used, and they took that mandate seriously. But in the absence of clear guidance and shared norms, they filled in the gap in ways that didn’t always move the business forward.  

They built models that didn’t match the decision cadence of the business. They communicated poorly, often burying insights in technical language or vague “demystification” efforts. They tried to “educate their stakeholders” with machine learning diagrams, confusion matrices, and model diagnostics that no one understood. 

And with bravado, they proclaimed the five worst words in data science: “Our model speaks for itself.”

This dynamic came into sharp focus for me a few years ago, while I was leading a firm-wide data training program for a Fortune 100 company. Just before the big program kickoff, one of the business unit presidents pulled me aside.

“I’m glad you’re here,” he said. “The data team is pissing me off. They’re smart, but they’re always trying to tell my team what to do.”

The problem wasn’t the data, but the dynamic. The data team hadn’t positioned themselves as partners in decision-making but instead were perceived as pushing conclusions the business hadn’t asked for. 

In trying to earn a seat at the table, some data teams forgot why they were invited in the first place: to inform, not override, decisions. We had asked for partnership.  Instead, we got parallel paths.

If this conversation resonates with you, join me at the CDO Magazine Chicago Leadership Summit on September 10, 2025, where I’ll be diving deeper into these themes during my breakout session: “Who’s Leading Whom? The Evolving Relationship Between Business and Data Teams.” We’ll explore how organizations can bridge the gap between analytics and real-world impact and ensure both sides of the business are truly aligned.

From parallel paths to common purpose

Today’s best organizations bring both perspectives together. We don’t need businesspeople to become amateur data scientists, and we don’t need data professionals to become proxy decision-makers. We need organizations built around a common purpose: to identify the most important problems and generate the best available evidence to make the best possible decisions.

That’s always been the goal. But for the first time, I believe we’re moving toward a more sustainable and steady-state balance, where both sides understand not just what they do, but how they contribute to something larger.

Today, I teach business and data leaders that decisions and fluency with evidence are the new center of gravity. It’s a familiar theme, but “evidence fluency” reframes the mandate in sharper terms: to think critically about trade-offs, uncertainty, actionability, and impact.

One of the clearest examples I use in my teaching comes from a multi-billion-dollar grocery chain that hired a data science consultancy to reduce losses from two persistent problems: stock-outs and spoilage. In pursuit of tighter operations, the consultancy built a neural network to forecast category-level demand.

The pilot results were remarkable. The model improved inventory-ordering accuracy by 42%, a projected $104 million in annual savings, at a cost of just $16 million to implement. A 6.5x ROI.

And yet… the board walked away. They declined to roll the model out more broadly because of three problems: none of them technical, and all of them familiar.

First, the business couldn’t reasonably act on the output. A predictive model might tell you to adjust your avocado order for tomorrow, but that doesn’t mean your supplier or your store infrastructure can adapt that quickly.

Second, the methodology created discomfort. A neural network’s non-transparent inner logic is difficult to explain, let alone defend at the board level. When business leaders don’t understand how something works, they often choose not to trust it.

Third, there had been data quality issues early on. Though the errors were minor and quickly corrected, they left a mark. Once a business team loses trust in the data, even temporarily, that trust is hard to earn back.

The result? A well-designed, high-ROI solution was shelved. Tens of millions of dollars in annual savings were left on the table. Not because the model was wrong, but because the evidence didn’t earn trust, and the business wasn’t built to act on it.

Becoming truly evidence-driven

​​When organizations commit to becoming evidence-driven, the dynamic between data and business teams changes. They stop working in sequence and start working in partnership. They don’t pass off insights, but shape and apply evidence together. Great data leaders become stewards of evidence, ensuring that it’s accurate, trusted, and used.

In an evidence-driven enterprise, decisions are treated as strategic assets. The path to OKRs and KPIs is through the quality of those decisions. At the end of the day, every competitor can copy your tech, hire your people, or match your capital. But your decisions define you: where to invest, what to build, and when to walk away.

If decisions are your greatest asset, then how you build and support them is what matters most. And that work depends on three things:

1. A shared commitment to the burden of proof

The foundation of any effective business-data partnership is a shared understanding of what actually counts as evidence.

Without it, teams risk offering solutions that don’t stand up to scrutiny, don’t translate into action, or don’t move the business forward. A shared burden of proof makes sure that everyone is working from the same assumptions about what’s convincing and credible.

This shared commitment is the foundation that allows the organization to decide with clarity and confidence. Business leaders buy in because they can defend the decisions they make. Data leaders focus on and invest in the right things because they know what’s truly needed to support those decisions.

2. Evidence ownership, not insight handoff

In many organizations, data work looks like a relay race: the data team builds something, hands it off, and steps back. But that model can break down. In high-functioning teams, no one gets to say, “I’ve done my part,” and walk away. If the business never uses the insight, the work wasn’t finished. If the model produces outputs that no one can operationalize, it wasn’t the right model.

Real partnership means co-owning the case for action. Business leaders shape the questions and stay close enough to the evidence to believe in it or reject it. Data leaders relentlessly stress-test whether the outputs will and should drive real decisions and lead to real results.  

Evidence, in the ideal case, is a shared asset. And when both sides take ownership, the work effectively drives important decisions instead of just filling decks.

3. Commitment to smarter bets, not better insights

Too often, data work is framed as a search for the right answer. But in business, there are no right answers in advance. There’s only uncertainty and bets. The real value of data lies in making smarter bets with clearer logic, stronger evidence, and better payoffs.

And if smarter bets are the goal, then ROI is the scoreboard, both for business outcomes and data investments as well. Measuring return should be a shared responsibility. When both sides are accountable for whether a decision delivers sufficient value, they plan better, build smarter, and stay focused on what matters most.

And when decisions consistently show measurable impact, a virtuous cycle takes hold: trust grows, and the smart use of data reinforces itself.  

Defined roles, shared stakes

The idea of treating data as evidence isn’t new or controversial. But what’s still emerging, and what will define the next phase, is how intentionally we share the work of producing and acting on it.

This is about maturing, not blurring, the relationship between business and data.  Mature teams don’t try to do each other’s jobs; they build trust by understanding their own roles and holding themselves accountable for the decisions they shape together. Business leaders stay close enough to the work to understand it and act with confidence. Data leaders care as much – if not more – about downstream impact than analytic soundness.

An evidence-driven enterprise isn’t one where evidence merely exists. It’s one where both sides take responsibility for making it matter.

Found this article insightful?

Don’t miss the chance to hear more from professor Joel Shapiro at the CDO Magazine Chicago Leadership Summit on September 10. In his breakout session, “Who’s Leading Whom? The Evolving Relationship Between Business and Data Teams,” Shapiro will dive deeper into how organizations can bridge the gap between business and data teams. With years of experience guiding both sides, he will share practical insights on how to become truly evidence-driven, build trust, and ensure both business and data leaders are aligned and accountable.

It’s a session you won’t want to miss if you’re looking to unlock real, lasting value from your data.