CIOs: Get to Know Your Algorithms

As a big sucker for the phrase “recommended for you,” I am quite possibly the ideal online shopper. (My other trigger words are “free shipping” and “special clearance.”) Given my profession and the number of CIOs I interview, I’m also well aware that all manner of fascinating data analysis is bubbling away in the database cauldrons of my favorite shopping sites.

But algorithms? The very word strikes an uneasy chord in the heart of anyone who is, to put it kindly, mathematically challenged. When my editors first proposed the story “CIOs Have to Learn the New Math of Analytics,” I wondered what on earth we might say about these undercover math equations that merited attention from our CIO readers.

Quite a lot, as it turned out.

The almighty algorithm is the fuel for today’s data-driven businesses. They stoke the data engines that recommend purchases, trade stocks, predict crime, spot medical conditions, monitor sleep apnea, find dating partners, calculate driving routes and so much more. “These math equations,” writes Managing Editor Kim S. Nash, “may someday run our lives.”

In the wrong application, they may someday ruin lives, as well.

The fascinating story that Nash unearthed will show you exactly why CIOs need to develop what one expert called “algorithmic accountability.” Understanding a few of the basic concepts used to build algorithms (no math degree required) could save your company from legal liability, poor business results or bad press.

Uber learned that the hard way. The taxi disrupter uses an algorithm that jacks up prices when demand is high. That nifty set of equations was only doing its job last December in Sydney, Australia, when it quadrupled the rates for terrified people fleeing a café under siege by an Iranian gunman. Uber apologized afterwards and refunded fares.

Our story also delves into the business upsides of algorithms at companies like UPS, H&R Block and eHarmony. And it makes the strongest case of all for IT leaders to think more deeply about business ethics and practices, rather than merely creating the data-handling systems that feed the algorithms. CIOs should understand the way data is being defined, monitored and manipulated in these systems.

My recommendation for you? Read this story and then go have a friendly chat with your best data engineer. Then you’ll understand why the almighty algorithm merits your attention.