In today’s data-saturated business environment, it’s almost a given that any presentation will include tables, charts, and metrics. Data has become the currency of credibility—a prerequisite for appearing thorough and informed. However, while this abundance of data holds immense potential, it also introduces significant challenges: the insights presented are often at risk of being superficial, irrelevant, or even misleading.

Consider these sobering statistics: decisions drive 95% of company performance, yet decision-makers fail to use best practices 98% of the time.¹ Moreover, a McKinsey survey revealed that poor decision-making in a typical Fortune 500 company wastes 530,000 days of managerial time annually—equivalent to $250 million in wages. These inefficiencies ripple through organizations, with employees spending an average of 47 days each year addressing the fallout from bad decisions, including rework, delays, and missed opportunities.²

Clearly, the problem isn’t a lack of data—it’s how we use it. This brings us to a troubling phenomenon: the data veneer.

The Data Veneer

Imagine sitting in a typical strategy meeting. A presenter clicks through their slides, which feature slickly designed charts and tables showcasing revenue trends, customer behaviors, or market shares. At first glance, it feels like the decisions being discussed are data-driven. But as you dig into the numbers and the narratives, a troubling realization emerges:

• The data lacks context—no detailed explanation of why these numbers matter to the decisions at hand (the decision to be made can be poorly defined, too).
• The insights are generic—statements like “leads are up 10%” or “page views are down 2%” that don’t clarify a root cause or provide actionable recommendations.
• When decisions are made, the connection to the presented data is tenuous at best, as if the data was included more to check a box than to drive clarity or direction.

This phenomenon—the data veneer—occurs when data serves as a decorative element to bolster credibility rather than as a genuine decision-making tool. While this veneer might create the impression of being data-driven, it rarely delivers the substance needed to guide meaningful action.

Yet, these challenges are not inevitable. The true test of data-driven decision-making lies not in the presence of numbers on a slide, but in the clear and logical bridge between those numbers, the insights they generate, and the decisions they guide.

The Slippery Slope of Confirmation Bias

The risk doesn’t stop at superficiality. The next step on this slippery slope is the misuse of data to support predetermined narratives or goals. Here’s an overly-simplified example of how it might play out (using broad brushstrokes to illustrate a point):

A marketing team receives an industry report claiming that a specific social media platform is the key driver of customer engagement and brand loyalty. The report highlights broad industry trends and compelling case studies from unrelated companies, aligning perfectly with leadership’s desire to increase social media investment. In response, the team combs through their internal metrics, cherry-picking data that seemingly supports the report’s claims—such as rising follower counts or increased post impressions. However, critical data showing low conversion rates or minimal impact on revenue is omitted. The resulting narrative is persuasive but fails to reflect the full picture, leading to a decision that prioritizes perception over actual business value.

I call this data shoehorning—a type of narrative engineering that involves the deliberate construction of a story using data, even when the supposed connections are tenuous or misleading. This approach often stems from cognitive biases like confirmation bias—seeking out evidence that supports a desired conclusion while ignoring data that might contradict it. Worse, it creates the impression of rigor while fostering false confidence in the resulting decisions.

It’s important to remember: while we should strive to tell a clear and compelling story with data, the data must lead the story—not the other way around.

The Consequences of a Flawed Approach

When data is used this way—as either a veneer or a tool for supporting preexisting beliefs—the risks are significant:

1. False Confidence: misleading narratives can lead teams to make decisions they wouldn’t have made with a more honest assessment of the data.
2. Missed Opportunities: genuine insights that could challenge assumptions or reveal untapped opportunities are overlooked.
3. Erosion of Trust: teams and stakeholders may start to question the integrity of the data science process when they realize data is being used primarily for unsubstantiated storytelling.

How to Reclaim Data’s True Purpose

To avoid falling into these traps, businesses must shift their mindset from simply including data to genuinely applying it:

1. Focus on Discovery, Not Validation: treat data analysis as an exploration aimed at uncovering truth, even if it challenges preconceived notions.
2. Ensure Traceability: insights should be directly traceable to their source data, with clear explanations of how they were derived, including any underlying assumptions or constraints that could effect interpretation.
3. Cultivate Constructive Skepticism: encourage teams to critically evaluate narratives, especially when the data aligns too neatly with desired conclusions.
4. Foster “Decision-Driven” Data Practices: start with the decision that needs to be made, and work backward to determine what data and analysis are necessary to inform it.

By adopting these principles, businesses can restore the integrity of a sometimes uninspired data science process and unlock its true potential for driving better decisions.

The Perfect Partnership: Decision Sciences And GenAI

This is where decision sciences truly shine. By blending rigorous analytical methods with a focus on direct decision-support to help choose between options — shifting focus from the science of inputs (data and analysis) to the science of outputs (recommendations and decisions)³ — decision sciences provide the framework needed to move beyond superficial data practices. They ensure that data not only informs but also empowers organizations to make smarter, more impactful decisions.

As we look toward 2025, Generative Artificial Intelligence (GenAI) is set to revolutionize data science by enhancing efficiency and enabling professionals to focus more on strategic decision-making. Traditionally, data scientists have spent a significant portion of their time—estimated at 70-80%—on data wrangling and transformations. GenAI streamlines these processes, reducing the time required for complex data preparations and enabling professionals to shift their focus to more value-driven activities.

This newfound efficiency allows data scientists to allocate more time to exploratory analysis, model building, scenario analysis, optimization, and interpretation of results. With GenAI acting as a heuristic guide, data science professionals can explore a wider array of methods and spend more time interpreting analyzed data. Decision-making happens through interpretation; looking at the possible outcomes, comparing alternatives, and choosing the best option. 

In doing so, Decision Sciences coupled with GenAI helps break through the data veneer by shifting the focus from superficial presentations to actionable insights tied to clear objectives. It bridges the gap between data abundance and strategic use, enabling Decision Scientists to deliver precise, forward-looking recommendations that resonate with stakeholders and drive visible outcomes.

A Call to Action for 2025

The proliferation of data and “insights” in business today has, at times, undermined the process of generating high-quality, decision-ready outputs. But 2025 offers an opportunity to reset. By embracing decision sciences as a rigorous discipline and leveraging advancements like GenAI, organizations can move beyond the data veneer and unlock the transformative potential of true data-driven decisions.

As reiterate the critical point made by Erik Larson in this Forbes article, it is important to consider that decision-making is both “the most important and most poorly managed business activity”. These two numbers tell the story:⁴

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

In addition, according to a 2017 Gartner report, 85% of big data projects fail⁵, while VentureBeat in 2019 highlighted that a staggering 87% of data science projects never reach production.⁶ Furthermore, Gartner’s 2019 prediction indicated that through 2022, only 20% of analytic insights are expected to deliver business outcomes.⁷All the above statistics underscore the significant challenges faced by organizations in harnessing the full potential of data science initiatives, and the fall-out for effective data-driven decision-making.

The question is no longer whether your business has enough data, but whether it has the right processes, tools, and mindset to harness that data for meaningful impact. Decision sciences provide a roadmap—one that transforms abstract insights into actionable strategies and measurable outcomes.

To lead in this new era, leadership must foster a culture that prioritizes clear objectives, measurable results, actionable recommendations. By taking this step, businesses won’t just keep up with the data revolution—they’ll lead it.

1. Can Market Researchers Save The Day By Becoming Decision Scientists? ↩︎
2. Three keys to faster, better decisions ↩︎
3. Since Data Scientists Don’t Do Business Decisions — Businesses Need Decision Scientists ↩︎
4. Can Market Researchers Save The Day By Becoming Decision Scientists?  ↩︎
5. Why Big Data Science & Data Analytics Projects Fail ↩︎
6. Why do 87% of data science projects never make it into production ↩︎
7. CIO: Transforming analytics into business impact ↩︎