Ditching Data Dogma

One of the most overlooked aspects in data proficiency is - nuance. Data professionals often grapple with a fundamental question: Is it better to be quick or right? This seemingly simple query unveils a complex landscape of trade-offs, strategies, and unconventional truths that shape the field of data science.

Pete Fishman, co-founder and CEO of Mozart Data, offers his perspective on this dilemma. With years of experience in data analytics across various industries, Fishman's insights challenge conventional wisdom and highlight the evolving nature of data-driven decision-making.

Slow and Wrong

Contrary to popular belief, the choice between speed and accuracy isn't binary. Fishman argues that effectiveness should be the primary goal:

"You probably need to be effective first, which is probably almost like the product of those two things. So you lose a lot of credibility if you are slow and you lose a lot of credibility if you're wrong."

This balanced approach recognizes that neither lightning-fast inaccuracies nor painfully slow precision serve businesses well. Instead, data professionals must strive for a sweet spot where timely insights meet reliable analysis.

The Myth of the Data Hero

Fishman notes that the "data hero" who sweeps in and saves a business with timely and eye opening data based conclusions simply doesn’t exist. The idealized scenario of a data analyst uncovering a groundbreaking insight that revolutionizes a company overnight is, in Fishman's words, a fantasy:

"This has literally never happened in my entire career, or I've never seen any version of this happen in my entire career. And I've worked in data for a very long time and I am the biggest proponent of data can change outcomes for companies."

Instead, the reality of data analytics is far more nuanced. It involves painting a richer picture of business operations, enabling faster reactions to changes, and supporting decision-makers rather than replacing them. This perspective shifts the focus from seeking silver-bullet solutions to fostering a culture of continuous improvement and data-informed decision-making.

Intuition vs. Data-Centricity

Contrary to the stereotype of data professionals as purely objective number-crunchers, Fishman argues that the best analysts often rely heavily on intuition:

"Actually most of the best data people have really strong priors and intuitions. And in general, if you're a Bayesian like me, a big part of having some sort of posterior beliefs after observing a bunch of data is what was the prior that went into it."

This balance between gut feeling and data-driven insights is crucial. While data should inform decisions, the ability to apply context, experience, and critical thinking to interpret that data is equally important. Ego can play a role here, as analysts must be confident enough to trust their intuitions but humble enough to let the data challenge their preconceptions.

AI Symbiosis 

As artificial intelligence and machine learning continue to advance, the role of human data analysts is evolving. Fishman draws the following parallel:

"I think we're in this sort of beautiful zone, just sort of our chess sense, where the highly skilled data person that can leverage AI effectively to make the parts rot, but also having really great instincts about when something is off or when something is wrong, I think we'll pay very huge dividends."

This analogy to chess, where human-AI collaboration proved superior to either humans or AI alone for a significant period, suggests that the future of data analytics may lie in symbiotic relationships between skilled analysts and AI tools.

The field is not about finding perfect answers or replacing human judgment with algorithms. Instead, it's about leveraging a combination of technical skills, critical thinking, intuition, and emerging technologies to drive better decision-making.

Success will likely come to those who can balance speed and accuracy, embrace the complexity of real-world data challenges, and cultivate a workforce that combines analytical rigor with creative problem-solving.

The journey toward data-driven decision-making is not a straight path to a digital utopia, but rather a continuous process of learning, adapting, and refining our approach to extracting value from information.