Twelve years ago — just over ten years into my career at the time —I wrote a blog titled, “3 New Metaphors to Inspire a Marketing Analytics Renaissance,” which, surprisingly, is still live today (broken links and all). Revisiting it has prompted me to reflect on how much my thinking has evolved. Looking back, I see that many of my original views still resonate, but my understanding of what it means to drive value through data has deepened, and (I like to think) I’ve grown into a more skilled and pragmatic data scientist along the way—one who understands that insights, no matter how brilliant, are meaningless if they don’t support decision-making.

“Data Helps, People Deliver”

In 2013, I wrote:

”The key to data-driven success lies is in how we use (or don’t) data to engage and motivate our teams. My motto: Data helps, people deliver.”

When I first wrote, “Data helps, people deliver,” it felt like a perfect encapsulation of the necessary dynamic between data and action. It also speaks to a healthy team atmosphere. A decade later, I still believe in its core message. Data provides direction (or it should), but people are the ones who ultimately create value by acting on it.

That said, I now understand that this connection isn’t automatic and cannot be taken for granted. Powerful insights are wasted without a willing audience. If data doesn’t actively drive decision-making—because decisions are the bedrock of “actionable” insights—then my motto above leaves something to be desired.

As I wrote back then:

“People give data value. That’s the truth. And the more valuable the data is perceived to be, the more actionable it becomes. Value and action are positively correlated. Quite literally, insight is in the eye of the beholder.”

Looking back, this observation foreshadowed a key principle of what I now know as Decision Science, a specialized discipline within data science dedicated to tightening the correlation between value and action. It acknowledges that while data may be objective the process of interpreting and acting on it is often surprising subjective, shaped by the context, perspective, and biases of the decision-maker.

This is the paradox of data science in business: even though data science strives for empirical precision, business environments are often political, influenced by competing agendas, cognitive biases, and organizational pressures (often, all at once). These factors can distort decision-making, undermining even the most robust analysis or well-designed models.

This is how I saw it at the time:

“Data represents opportunity. This opportunity should be a great mobilizer. As marketing teams, we should be proactively exploring and probing these opportunities, considering different scenarios, testing hypotheses – and using data as storytelling attributes!”

When I wrote this over a decade ago, I believed in the transformative power of data to inspire teams to act and innovate through exploration and discovery. I was glad to see that even then, I understood the importance of scenario analysis and hypothesis testing, for example, as essential tools for turning data into a “great mobilizer.” As I wrote, “data without action is overhead”—a truth that remains relevant today.

However, I’m less optimistic about the progress we’ve made. Scenario analysis, or “what-if” analysis, for example, remains a completely underutilized yet incredibly powerful decision-support method. Too often, however, the “insight industrial complex”—the tools, practices, and mindsets that prioritizes producing insights over applying them—dismisses these methods as “nice-to-haves” in the face of deadlines or other constraints.

Storytelling, while still essential, must also evolve. Stories should emerge from the data, not the other way around. Crafting narratives first and forcing data to fit them—what I call data shoehorning—distorts insights and undermines credibility. This can range from leaving critical gaps in an analysis for stakeholders to fill in with assumptions, to outright manipulation of data to suit a narrative, echoing Noam Chomsky’s concept of the “manufacturing of consent.” In Chomsky’s framework, public opinion is shaped through selective information to serve an agenda; in the world of insights, this translates to selectively framing data to push a predetermined conclusion, often at the expense of accuracy or trust.

As Erik Larson emphasized in a Forbes article, decision-making is both “the most important and most poorly managed business activity.” The numbers speak for themselves:

• Decisions drive 95% of company performance.
• Decision-makers fail to use best practices 98% of the time.

Although data-driven storytelling is not the only culprit behind these alarming statistics, getting it right depends on addressing the fundamentals of effective decision-making first. By contrast, when done well, data-driven storytelling preserves the integrity of insights, builds trust, and guides action. In other words, data-driven storytelling should never be open-ended. Instead, it should present clear options, quantify their anticipated impact, and provide the guidance necessary to turn insights into confident decisions—or, when confidence is unattainable, explain why a decision inherently lacks certainty… while remaining confident in that conclusion!

From Data Science to Decision Science

Perhaps the most important lesson I have learned over the last decade—as I have evolved from a data scientist to a self-identified decision scientist—is that it is not enough to produce insights, even great insights; we need processes and frameworks that prioritize the science of outputs (recommendations and decisions), over the science on inputs (data and analysis), to drive consistency and integrity.

Furthermore, we must play a critical role in the effective use on GenAI in the “new economy of insights,” where Decision Scientists will collaborate with GenAI tools to generate model outputs dynamically and in formats tailored to the specific needs of the stakeholders. The value of an insight isn’t just in its generation but in how it’s contextualized and applied.

Experience has taught me that this subtle shift in focus — from topic-driven analysis to decision-driven analysis and acknowledging the necessity of the interpretation of analyzed data (i.e. comparing alternatives, exploring outcomes, and determining the best course of action) to strengthen the correlation between data and action — is critical to ultimately bridging the gap between “data helps” and “people deliver.” We need decision science skill-sets and frameworks — not just analytical busy work, or “doorstop reporting” — to establish continuity between the two.

So, reflecting on my original motto, I see it as a starting point rather than an endpoint. “Data helps, people deliver” still resonates, but it now requires a more nuanced understanding:

• Data Helps: data’s value isn’t inherent—it comes from its ability to inform and guide decisions. This requires prescriptive, goal-aligned recommendations.
• People Deliver: people remain the critical link between data and action, but they need the right outputs to succeed—comparing alternatives, exploring outcomes, and determining the best course of action.

Ultimately, the connection between “data helps” and “people deliver” demands accountability to close the intelligence value gap—the chasm between what an organization currently knows and what it needs to know to make strategic decisions.

Keeping it Real

To be clear, Decision Sciences doesn’t eliminate the challenges of politics, competing agendas, or cognitive biases, but it acknowledges these hurdles and works to create transparency where decision “black holes” often exist. These black holes—where insights and accountability disappear—are avoidable with clear documentation of decisions and their intended outcomes. By defining a clear impact path upfront, organizations establish accountability, making it possible to track whether decisions were made (or not) and measure their effect on outcomes.

Insights without action are whispers in the wind.

And, as I have written very recently, at the end of the day the role of a Decision Scientist is to make insights usable—building the connective tissue between “data helps” and “people deliver”—by providing increasingly predictive and prescriptive recommendations.

Data as connective tissue metaphor (click to read)

In 2013, I wrote:

“This metaphor is rooted directly in biology. Connective tissue holds things together.  Its function is to support, strengthen and connect.  Is this not what we want from the “actionable insights” that are in such high demand within our marketing organization?  Sure, we all want our performance numbers to trend up and to the right. A given. But take care of the support, strengthen and connect part and performance will take care of itself. I promise.  In this sense, from my perspective, actionable insights are connective tissue.”

Looking Ahead

The lessons I’ve learned over the past decade have tempered my early optimism but deepened my appreciation for the complexity of turning data into action. Decision Sciences formalizes this process, ensuring that data leads to decisions, and decisions lead to measurable results.

The work is harder, and the challenges are greater, but the potential for impact has never been clearer. As I look ahead, I see my role not just as a provider of insights but as a guide—helping teams and organizations bridge the gap between data and decisions.

In this new era of GenAI, the role of Decision Scientists will become even more critical. As I explored in my last blog post, GenAI tools have the potential to revolutionize the speed and scalability of insights, but their outputs are only as valuable as the context and decision-making frameworks that shape them. Decision Scientists will be essential in ensuring these tools are applied effectively, playing the extended role of Curators, Validators and Strategists.

In the “new economy of insights”, the interplay between data, people, and GenAI will define success. Data helps by providing the foundation, people deliver by making decisions and driving action, and GenAI amplifies both—scaling insights, accelerating analysis, and tailoring outputs to specific stakeholder needs. But as powerful as GenAI is, it cannot replace (yet) the human ability to interpret, contextualize, and prioritize insights. Decision Scientists will be the linchpins in this new ecosystem, ensuring that data, people, and GenAI work seamlessly together to deliver outcomes that matter.

We might say now, as we look towards the future:

• Data Helps: data’s value isn’t inherent—it comes from its ability to inform and guide decisions. This requires prescriptive, goal-aligned recommendations.
• People Deliver: people remain the critical link between data and action, but they need the right outputs to succeed—comparing alternatives, exploring outcomes, and determining the best course of action.
• GenAI Amplifies: GenAI amplifies both data and people by accelerating insights, scaling analysis, and tailoring outputs dynamically to stakeholder needs. Its true power lies in collaboration with human expertise, ensuring that its outputs are contextualized, actionable, and aligned with strategic goals.

What are your thoughts on the evolving role of GenAI and Data Science in closing the intelligence value gap? I’d love to hear your perspective.