Lost in Translation: Bridging the Gap Between Data Complexity and Business Simplicity

Time and again, I come back to this picture to illustrate what generates business value. There is a HUGE disconnect between the data and the AI community’s “perception” of what business users want vs what business users really want. Every once in a while, data scientists will come up with some complex visualization depicting detailed statistics, thresholds, accuracies, and percentage errors, and every time, business users have the same puzzled expression on their faces. And business users almost always come back to the same question, “But what is the business value in this?”. The routine has become predictable, tragic, and almost comical.

To the utter disbelief of the Data and AI community, business users always complain “We don;t have the data or we don’t have the right insights”. Then, they never open the elegantly designed reports or visit the elaborate dashboards that are updated in real-time. In fact, based on experience, the utilization of complex analytical dashboards and reports is less than 20%, and that too during month/quarter ends when such data is needed for reporting instead of day-to-day decisioning. So what’s going on?

The crux of the issue lies in the fundamental differences in how value is perceived. For data scientists, value often equates to the intricacy of algorithms and the elegance of solutions. They revel in the sophistication of models that can process vast datasets and output precise metrics. In their world, complexity signifies capability, and the more granular the data, the better.

On the other side of the spectrum, business users gauge value through the lens of practicality, usability, reliability, timeliness, and actionability. They prioritize straightforward solutions that provide clear and reliable directions for important business decisions—much like a GPS navigation system that offers the most efficient route to a destination and tells the ETA without overwhelming the driver with the intricacies of the route’s calculation.

This disconnect leads to a paradoxical scenario where both parties are working towards the same goal—driving business value—yet are miles apart in their approach. The data scientists’ complex dashboards, brimming with real-time updates and intricate details, remain untouched while business users continue to demand more ‘data’ and ‘usable’ insights.

This ubiquitous narrative highlights the contrasting backgrounds of the two parties and a crucial gap in communication and understanding. Data scientists must recalibrate their focus towards simplicity and relevance. They need to ask, “What is the most burning business question, and which insight will offer the highest marginal utility?” and “How can we translate these complex data points into straightforward insights that drive business decisions?” Marginal utility is an important distinction because, depending upon the business maturity, what has the highest utility changes from business to business, region to region, and time to time. If you are lost in the middle of a dense jungle, knowing which way is the nearest village has the highest utility. If you are lost in a large city, you need a more ‘tangible’ landmark.

Business users, for their part, must articulate their needs more clearly, specifying the kind of data and insights that would be most beneficial for their strategic goals. In fact, I have been an advocate of more hands-on business involvement in data and AI initiatives. Business users are spoilt by the outsourcing mindset inherited from IT, where business analysts translate business requirements. However, it is also a matter of interest and maturity as many business users don’t see data and AI as part of their job description, which is a BIG mistake and could endanger your job and career in most jobs in the next 5-7 years.

In the meantime, the role of ‘Business Translators’ or ‘Value Navigators’ becomes indispensable in bridging this gap. These professionals can offer tactical support to business leaders by identifying opportunities, interpreting business objectives and domain constraints, and translating them into data science objectives. They also help data scientists design appropriate solutions and translate AI inferences into business decisions. They serve as the missing link that can turn complex models into intelligent decisions that business users can readily understand and act upon.

Ultimately, the solution lies in creating such ‘Decision Intelligence’ bridges between business opportunities and Data/AI assets and building commercial frameworks to translate between the two. It is critical to foster a collaborative environment where data scientists are deeply integrated into the business side, understanding the core objectives and strategies. Simultaneously, business users should be brought into the data science process, providing context and purpose to guide the technical work. Only by walking in tandem can both sides reach the shared destination of tangible business value, avoiding the detours of miscommunication and the dead ends of unmet expectations.