A friend works for one of the world’s largest online retailers. In any given day, transactions on their sites generate more than 1 billion gigabytes of new data. Algorithms mine that data constantly, and business unit leaders and the top executive teams make business decisions based upon the patterns they unearth. But there’s a nagging problem: a perception of fake reviews flooding the site. The noise in the system is that these reviews so complicate a customer’s evaluation and decision-making efforts that shoppers have started to go elsewhere. Yet the algorithm’s analysis is clear: less than 3% of reviews are fake. My friend recently took over a re-look at this problem, which prompted this discussion with the project manager:

“I understand,” my friend started, “that the analytics are consistent. But what’s odd is that I’ve personally experienced fake reviews. My partner has experienced fake reviews. Both our sets of parents have encountered fake reviews.”

The project manager was nodding. He, too, has personally experienced the fake review problem, as has nearly everyone he’s spoken with casually about his project.

“So what does that tell us?” my friend asked.

The project manager’s answer was near-immediate: “The algorithm is wrong. We have to look at the data a different way.”  And so their new undertaking has begun.

Why do I tell that story? Because the most technologically sophisticated, data- rich, analytics-capable organizations in the world have to learn anew, daily, how to best use oceans of data to extract the pearls of real value. For smaller organizations, for manufacturers, for board members and HR business partners, the question of how to make meaningful business decisions based upon data analytics is still very much a work in progress. The hype of big data and analytics is still largely unrealized. “More” and “big,” in this context, does not translate to clarity or usefulness.

Yet organizations are making headway, and in this issue we seek to spotlight a few examples and ask them to unpack the “how” of their process, their findings and what they see as the next phase or challenge-set for their own evolution as HR and business leaders. Contributors include a range of senior HR leaders, CHROs and their analytics teams, academics, technology/data professionals and board members.

As guides on this process, Judith Scimone, Senior Vice President, Chief Talent Officer of MetLife, and Claudy Jules, Ph.D., Partner of McKinsey & Company, graciously agreed to lend their expertise—which stem from their histories as HR leaders and organizational thinkers—to help curate a series of articles and questions to help each of us, in our own context, challenge our thinking and use of data to drive pragmatic insight and improve—change—fact-based decision-making. An overarching theme in all of this is that data is not truth—it’s data. The truth is hidden somewhere inside it, and what one chips away to reveal the truths is just as important as what one focuses upon or keeps.

Going back to that online retailer’s fake-review dilemma, one of its key challenges is essentially using data to plumb the veracity of sentiment. How do you know if an enthusiastic or poor review is real? Many of HR’s current data challenges face a similar curve. How do we know that our data on culture, on ESG impact broadly, on diversity in our top-to-bottom management and leader pipeline is real? In many cases, organizations have relied for years on analyses that produced the most attractive answer (only 3% fake reviews!) rather than the analyses that prompted the hardest, but most impactful discussions.

This issue presents leaders who are working hard to change that. Thank you for joining the discussion—and watch for additional articles on the SHRM Executive Network website to extend this conversation throughout Q1.   

David Reimer

Executive Editor