“If you do not know where you are going, every road will get you nowhere.” — Henry Kissinger
If you walk into a boardroom and ask, "What is the goal of our analytics function?", you will often hear answers like "to provide accurate reports," "to predict future sales," or "to build a data warehouse." These answers, while technically correct regarding activities, are fundamentally wrong regarding purpose.
The goal of analytics is not to produce charts, dashboards, or models. The goal of analytics is to reduce uncertainty in decision-making to achieve specific business objectives.
This is a simple and straightforward answer to the question, ‘what is the goal of analytics’, but it will take much more background to show how profound this specific goal is in the midst of competing answers to the same question.
Why is this goal more critical today than it was a decade ago?
First, the sheer volume of data has exploded. Consider the iconic 1994 photograph of Bill Gates suspended atop a massive tower of paper to demonstrate the efficiency of a single CD-ROM. That disc held roughly 700 megabytes—a staggering amount at the time, equivalent to some 330,000 sheets of paper. Today, that capacity feels quaint. The average internet user now generates roughly 143 gigabytes of data traffic per day—equivalent to over 200 of those CD-ROMs. To use the Gates analogy, if we printed that daily data out, every single person would generate enough paper to build a tower 6.8 kilometers high—eight times taller than the Burj Khalifa—every single morning.
We have moved from managing gigabytes to grappling with zettabytes, with global data creation projected to exceed 180 zettabytes by 2025. This volume is so immense that data is now referred to as the ‘exhaust’ of the information age—an unavoidable, continuous byproduct of digital existence. Somewhere in that exhaust is critical information that will determine the long term success of your business, and you are in a race with your competitors to find those insights.
Second, the nature of data has shifted. Ten years ago, businesses looked primarily at internal transaction logs. This was clean, predictable rows of data like "Item A sold for $10." Today, we are grappling with the data about data or metadata.
Consider a single Tweet. To the user, it is just 280 characters of text. But to the analytics engine, that single tweet generates a JSON object containing over 150 metadata fields: the exact GPS coordinates of the user, the device type (iPhone vs. Android), the timestamp down to the millisecond, the user's account creation date, their follower graph, and the engagement velocity (retweets per second).
The same applies to finance. A decade ago, a bank transaction was a simple record: "$50 withdrawal." Today, a digital payment generates a rich log of "behavioral exhaust": the IP address of the transaction, the geolocated merchant ID, the time spent on the checkout page before clicking "pay," and even the biometric data of the fingerprint used to authorize the sale. Anyone of these data points can be used to improve the quality or security of the transaction itself e.g. automatic fraud alerts for transactions coming from a new device or unrecognized location.
Furthermore, the shift to cloud computing has exponentially amplified this logging. In an on-premise server room, you might log basic access attempts. In a modern cloud environment (like AWS or Azure), every microservice, every API call, and every load balancer event generates a continuous stream of logs. This isn't just data; it is a high-velocity torrent of context. If you do not harness this to reduce uncertainty, your competitor will.
Today, businesses must evaluate how to integrate external, high-velocity streams: social media sentiment (like Facebook, TikTok do), real-time geospatial sensor data (like Uber, Waze do), and granular consumption preferences (like Spotify, Netflix do) into their products and services, or into their decisions about how to run their businesses. The average consumer has grown to expect more from their products such that a failure to deliver rich insights from data is seen as being under-performing, even when you are otherwise delivering a high quality product.
Finally, the infrastructure to process this has become democratized; powerful algorithms and cloud storage are now cost-effective. AI is being integrated into platforms to make them easy and intuitive for all users, even those without extensive data analytics training. If you do not harness this to reduce uncertainty, your competitor will. While you hesitate to gain a better understanding of how these data sources are influencing your market and your business, your competitors are racing ahead to do so.
This also leads to questions about the value of the data analytics products within organizations today, a subject we will spend a full chapter on in chapter three. If a report took two months to put together, or a dashboard is beautiful, accurate, and real-time, but does not help a manager make a decision that moves the needle on a business objective, it is a nonperforming waste. To understand this, we must look at analytics through two distinct lenses: the mandatory requirements of the regulatory environment, and the strategic competitive ready choices of the modern manager.