Nova: Imagine two companies in the same industry. One makes decisions based on gut feeling, hierarchy, and the loudest voice in the room. The other builds every strategic move on hard evidence and rigorous analysis. Research shows the second type is, on average, five to ten percent more productive, commands a higher market value, and delivers fatter returns to shareholders. That gap is not luck. It's methodology. And that's exactly what Claus Grand Bang's book, Data-Driven Decision-Making for Business, unpacks from cover to cover.

Nova: : Five to ten percent productivity boost just from being data-driven? That feels almost too tidy. What's the real story behind that number?

Nova: The number comes from a well-known study by Erik Brynjolfsson and his colleagues at MIT, who found that firms adopting data-driven decision-making outperform their peers across multiple metrics. Grand Bang builds his entire book around this premise. But here's the twist: his core argument is not that data itself has value. His central message is that value does not come from data, but from acting on data.

Nova: : So the data sitting in a warehouse somewhere is basically worthless until someone does something with it.

Nova: Exactly. And Grand Bang is uniquely positioned to make that argument. He spent a decade building data-driven companies in the business world, then a decade teaching applied data analytics in academia across Europe. He even created one of the first applied data analytics degrees in Europe. Now he's Head of Data and IT at a global biotech firm. He's lived both sides of the equation.

Nova: : That blend of academic rigor and battlefield experience sounds rare. So what's the roadmap he lays out in this book?

Nova: It's a ten-chapter journey across the entire landscape of data-driven decision-making. Strategy, culture, analytics, infrastructure, ethics, and even a forward-looking chapter on generative AI. Published by Routledge in 2024, it's designed for students and practitioners alike. And he peppers the whole thing with real case studies: Saxo Bank, the Bangladeshi fintech bKash, BBVA. These are not hypotheticals.

Nova: : I'm already curious about what a Bangladeshi fintech company can teach us about data-driven decisions. Let's get into it.

Nova: Most books about data start with technology. Grand Bang starts with decision-making itself. He introduces something called the DECAS model, which stands for Decision, Evidence, Choice, Action, and Sense-making. It's his way of integrating classical decision theory with modern data analytics.

Nova: : Break that down for me. What does each piece mean?

Nova: Decision is the framing: what are we actually trying to decide? Evidence is where data comes in. You gather relevant information. Choice is evaluating alternatives. Action is the execution. And sense-making is the feedback loop: did it work? The model forces you to treat decision-making as a structured process, not a gut reaction.

Nova: : So it's almost like the scientific method applied to business decisions. Hypothesis, experiment, result.

Nova: That's exactly right, and Grand Bang makes that connection explicit. He draws on Daniel Kahneman's prospect theory and decades of behavioral economics research. The whole point is to counteract what he calls systematic patterns of deviation from rationality. Basically, our brains are wired to take shortcuts, and those shortcuts cost money.

Nova: : But here's what I'm wondering: if the model is so powerful, why don't more companies use it?

Nova: Grand Bang addresses that head-on. He argues that most organizations have what he calls a decision-making gap. They collect enormous amounts of data, but the data never actually reaches the decision table. There's a chasm between the analytics team and the executives. His book is essentially a bridge-building manual.

Nova: : Speaking of bridges, how does he connect data strategy to business strategy? Because I can imagine a scenario where a company has a brilliant data team working on problems that don't actually matter to the bottom line.

Nova: That's Chapter Two of the book, and it's one of his strongest contributions. He uses Porter's generic strategies and blue ocean strategy frameworks, but maps them directly to data initiatives. The idea is that your data strategy should not exist in a vacuum. It should be a direct extension of your competitive positioning.

Nova: : Give me a concrete example.

Nova: He talks about BBVA, the Spanish bank. They didn't just say let's get better at data. They tied data practices directly to employee incentives. If you wanted to advance at BBVA, you had to demonstrate data literacy. That alignment between organizational incentives and data behavior is what Grand Bang calls the sweet spot.

Nova: : So it's not just about having data, it's about making data matter to people's careers.

Nova: Precisely. And that leads us directly to the next big theme in the book: data culture.

Nova: Chapter Four of Grand Bang's book tackles what might be the hardest piece of the puzzle: data culture. He argues you can buy the best technology stack in the world, but if your people don't trust data or don't know how to use it, you've built a Ferrari with no one who knows how to drive.

Nova: : That feels painfully true. I've been in meetings where someone presents a beautiful dashboard and everyone nods politely, then goes back to doing whatever they were doing before.

Nova: Grand Bang calls that the data theater problem. Organizations that perform being data-driven without actually being data-driven. His solution involves four interlocking elements: leadership intervention, data literacy programs, behavioral incentives, and transparent access to data.

Nova: : Leadership intervention sounds like code for the CEO needs to care about this.

Nova: It is, and he's unapologetic about it. He writes that without active and visible leadership commitment, data culture initiatives wither. He uses Saxo Bank as a case study. Their leadership made a deliberate decision to push transparency across the organization. Data wasn't hoarded in corner offices. It was made visible and accessible.

Nova: : But here's the tension: making data transparent sounds great until someone uses that transparency to second-guess every decision a manager makes. Doesn't that create fear?

Nova: Grand Bang acknowledges that. He frames it as a psychological shift from data as a weapon to data as a tool. If data is used to punish people, the culture will never take root. If data is used to learn and improve, people eventually embrace it. He references the concept of psychological safety, which is well-established in management research.

Nova: : What about data literacy? Because I know plenty of smart people who freeze up the moment you put a spreadsheet in front of them.

Nova: That's the literacy gap he dedicates significant attention to. He argues that data literacy is not about turning everyone into data scientists. It's about ensuring that everyone in the organization can ask intelligent questions of data, interpret basic visualizations, and understand the difference between correlation and causation.

Nova: : Correlation versus causation: the eternal trap.

Nova: He devotes space to that exact problem. He notes that one of the most common errors in data-driven organizations is confusing the two. Just because ice cream sales and drowning incidents both rise in summer doesn't mean ice cream causes drowning. That sounds obvious, but in business contexts, the mistake gets made constantly.

Nova: : So beyond mindset, what does he say about the actual mechanics of working with data? Where does it come from, and how do you manage it?

Nova: Grand Bang dedicates Chapter Three to something he calls data products, and Chapter Seven to the analytics trinity: descriptive, predictive, and prescriptive analytics. What ties them together is a methodology called CRISP-DM, which stands for Cross-Industry Standard Process for Data Mining.

Nova: : That sounds technical. Is CRISP-DM something only data scientists need to care about?

Nova: Grand Bang's argument is the opposite. He says businesspeople need to understand CRISP-DM because it's the bridge between a business problem and a data solution. The six phases are business understanding, data understanding, data preparation, modeling, evaluation, and deployment. Notice the first phase is not coding a model. It's understanding the business.

Nova: : So it's back to that alignment you mentioned earlier. Start with the business question, not the data.

Nova: Exactly. And he pairs CRISP-DM with what he calls the Data Product Canvas, which is a framework inspired by design thinking. It forces you to articulate who the user of a data product is, what problem it solves, what data it needs, and how success will be measured. It keeps data teams from disappearing into technical rabbit holes.

Nova: : What's an example of a good data product versus a bad one?

Nova: A bad data product is a dashboard with fifty metrics that nobody asked for, updated daily, and ignored by everyone. A good data product solves a specific, recurring decision. Like a churn prediction model that flags which customers are at risk of leaving and triggers a retention action. Grand Bang uses bKash, the mobile financial services company in Bangladesh, to illustrate this. They used data products to serve millions of unbanked customers by making credit decisions based on mobile usage patterns.

Nova: : That's fascinating. Traditional banks would look at credit history, but these customers had no credit history, so bKash had to find entirely new data signals.

Nova: Right. And that gets at the three types of analytics Grand Bang covers. Descriptive tells you what happened: how many transactions, what volume. Predictive tells you what will happen: which customer is likely to default. Prescriptive tells you what to do about it: offer a smaller loan with a different repayment schedule.

Nova: : Most companies I know are still stuck at the descriptive stage. They have reports showing what happened last quarter, but they're not predicting or prescribing.

Nova: Grand Bang calls that the analytics maturity curve, and he's candid that most organizations are lower on it than they think. But he also argues that you don't need to jump straight to artificial intelligence. Even basic descriptive analytics, done well and acted upon consistently, can generate enormous value.

Nova: : Speaking of artificial intelligence, the book has a whole chapter on generative AI. What's his take?

Nova: Chapter Nine on data ethics and Chapter Ten on generative AI are where Grand Bang gets really forward-looking. He argues that data-driven decision-making without an ethical framework is not just risky. It's unsustainable. He draws on regulations like the European GDPR and the California CCPA to ground the discussion in real legal obligations.

Nova: : But here's the question I always have about data ethics: everyone agrees it's important, but when the pressure is on to deliver results, doesn't ethics become an afterthought?

Nova: Grand Bang anticipates that. He uses governance frameworks like COBIT and the DGI framework to make ethics operational, not aspirational. The RACI matrix is one of his key tools: for every data decision, you identify who is Responsible, Accountable, Consulted, and Informed. Ethics doesn't get lost because someone is explicitly accountable for it.

Nova: : So it's not about having a mission statement that says we care about ethics. It's about assigning a real person whose job depends on it.

Nova: Exactly. And he extends this to the conversation about generative AI, which is his final chapter. His position is nuanced. He sees enormous potential for generative AI to accelerate decision-making: summarizing vast datasets, generating scenarios, even drafting decision memos. But he warns that generative AI amplifies existing biases if the underlying data is flawed.

Nova: : The garbage in, garbage out problem on steroids.

Nova: That's essentially his argument. He also raises a concern that's less talked about: the black box problem. If a generative AI model recommends a course of action, but nobody can explain why, what happens to organizational accountability? Who owns a decision made by an algorithm?

Nova: : That's the kind of question that keeps general counsels up at night.

Nova: And regulators too. Grand Bang doesn't offer easy answers here. He's honest that the generative AI chapter is more about framing the right questions than delivering final solutions. The technology is evolving too fast for anyone to have the last word.

Nova: : What I appreciate about that approach is the intellectual humility. Too many business books promise a silver bullet, and he's saying: here's the frontier, here's what we know, and here's what we still need to figure out.

Nova: That humility runs through the whole book. Grand Bang is not selling a fantasy of perfect, frictionless data-driven decision-making. He's selling a discipline. A way of thinking. A set of frameworks that, applied consistently over time, shift the odds in your favor. And he's got the research, the case studies, and the personal experience to back it up.

Nova: So let's bring it all together. Claus Grand Bang's Data-Driven Decision-Making for Business is built on a deceptively simple premise: value does not come from data, but from acting on data. Everything else flows from that insight.

Nova: : And he gives us a roadmap to act. The DECAS model to structure decisions. Data strategy aligned with business strategy. A culture built on literacy, leadership, and psychological safety. Practical frameworks like CRISP-DM and the Data Product Canvas to turn raw data into real products. And guardrails around ethics and generative AI to make sure we're building something sustainable.

Nova: What I find most compelling is how he bridges two worlds that often don't talk to each other: the world of classical decision theory with its deep understanding of cognitive biases, and the world of modern data engineering with its dashboards and machine learning pipelines.

Nova: : The cognitive bias piece really stuck with me. The idea that even with all the data in the world, our brains are still working against us. Confirmation bias, anchoring, availability bias. Grand Bang doesn't pretend data makes those disappear. He argues data gives us a fighting chance against them.

Nova: And the case studies bring it to life. Saxo Bank's transparency experiment. bKash creating financial access for millions of unbanked people using mobile data. BBVA tying career progression to data literacy. These are not abstract theories.

Nova: : So what's the one thing you'd want a listener to take away from this conversation?

Nova: I'd say it's this: being data-driven is not a technology project, it's a decision project. You don't start by buying tools. You start by asking what decisions matter most to your organization, and then you work backward to the data, the analysis, and the culture you need to make those decisions better.

Nova: : And then you act. Because the data sitting on a server somewhere is just potential energy. Grand Bang's whole message is about converting potential into kinetic. Data into value.

Nova: Beautifully put. If you're a business leader, a student, or anyone who participates in organizational decisions and you want to move beyond gut instinct, this book deserves a spot on your shelf. It's published by Routledge, available now, and it reads like a conversation with someone who has been in the trenches and knows what actually works.

Nova: : This is Aibrary. Congratulations on your growth!