In 2005, Urchin Software Inc. was just a fledgling startup in the Little Italy neighborhood of San Diego. Couched in a 1900s-era furniture warehouse, Urchin’s band of twenty-somethings were on a mission that had yet to spread to the rest of the world: helping customers to track online visitor traffic.
That April, Google bought Urchin’s software for an undisclosed sum and began building what is now known as Google Analytics.
At the time, the data collection behind Urchin was revolutionary. Today, it’s so universal we take data collection for granted. Systems companies rely on to do business — CRMs, marketing automation platforms, ERPs, mobile apps, and other digital systems — also collect and store data.
Even on an individual level, data collection has become so popular it’s given birth to the "quantified self" — the perpetual logging of our own lives for the purported end of self-improvement. Wearables like Fitbits and Apples Watches collect data about how we walk, sleep, and eat. Meanwhile, ordinary devices — from cars and refrigerators to voice assistants and smart TVs — show how data-driven performance has spilled into the home. While employees track views and installs, everyday samaritans monitor skin conductance and chart moods.
This ubiquitous obsession with data shows little signs of letting up. By 2020, Gartner predicts that 10% of organizations will devote a team expressly for productizing and commercializing their information assets. In 2012, just 12% of firms had appointed a chief data officer (CDO). By 2018, that percentage has grown to 63%. In tandem, businesses are also creating product roles to mine information assets. The idea being, if you parlay operational, commercial, and public data into more consumable and marketable forms, you open the door to more profitable opportunities.
Data has thus become a two-for-the-price-of-one asset. On the one hand, data can fan the flame of innovation, differentiating a businesses among a crowded, competitive marketplace. On the other, analysts can reflect upon history and find patterns — in their campaigns, advertising returns, conversion rates, operational inefficiencies, customer preferences, and more — to correct mistakes. With data, organizations can also forecast whether they should increase an investment in a certain product next quarter, and by what percentage. Analysis of why an email marketing campaign failed might indicate that the campaign wasn’t mobile-friendly and thus that you should create more responsive designs for future campaigns. In short, data can thus be a strategic asset that powers major decisions.
But collecting data is not without its costs and businesses should be wary not to rely on it for every decision. Those tasked with compiling and processing metrics must spend a considerable amount of time on data prep before it’s ready for analysis. In addition, data capture in the form of surveys, forms, and promotional emails can irk customers who are simply looking to sign up, get help from a customer service agent, or go about their daily life. Not to mention, as organizations collect more data, they can introduce novel security threats that can endanger property, identity, health, and even national security.
Clearly, if we are to prize data’s value, we must acknowledge and respect its risks. Although tempting, managers shouldn't totally supplant human judgment and real-world experience with statistics, since the latter may also mislead or paint a reductionist picture of the truth.
Businesses must also be aware that rewarding or penalizing employee performance based on data actually tends to lower performance outcomes. And when the stakes are high (e.g. keeping one’s job, getting a raise), workers are more inclined to satisfy those measures at the expense of other, more important organizational goals (innovation, risk-taking, experimentation, creativity).
With all this mind, being data-driven is still a good strategy. But you should use your data as you would any precious resource: prudently and within reason. Ask whether every planning meeting should begin with data. Don’t fail to make another decision just because you haven’t measured a past decision. Because if data is, in fact, an asset, such as a car or house, you should treat it accordingly. Use the right data. Plan how you use your data carefully. Build your data architecture with models that predict and optimize business outcomes. Maintain your data integrity over time, and use data without totally depending on its value for every single idea and pivot. In short, revere data while taking its value with a grain of salt. For as stewards of data, we should remember that just because something is hard to measure doesn't mean it isn’t real or important, that we also have a duty not to let short-term profits jeopardize our long-range goals. ♦