WHY DATA SCIENCE?
Let’s start by getting something straight: I have always hated statistics. Ever since 3rd grade, when I realized I didn’t even know how to say the word correctly (I called it Stat-Sticks), I have had a contentious relationship with the subject, especially during my time as a math major. I fell in love with math in the first place because of its elegance. For anyone who loves a good puzzle, perfectly factoring a binomial can sometimes feel as satisfying as pushing the final piece into a complex jigsaw. I loved all of mathematics…until I met Stats.
Statistics are dirty. They’re messy, complicated, annoying, unclear, and fiddly. On the first day of Stats 100, our teacher opened with the famous Benjamin Disraeli quote: “There are three kinds of lies: lies, damn lies, and statistics”. Ironically, Disraeli never said that, Mark Twain did, so the quote is a lie in and of itself. I left that class believing Statistics is “Math for Liars.”
Oh how naive I was.
MANAGEMENT 101
I have always been a problem solver, and so it was natural that I would eventually fall into management. Due to the unfortunate timing of my graduation in 2009, my first two months in real life were lost in the quagmire of unemployment until I found a position at a nearby hair salon. What did I know about hair at the time? Well, I knew it was hiring, so that was enough for me.
The salon industry taught me more about real life in 9 months than college did in 4 years. Over the next half decade, I moved from desk person to General Manager of four salons using the power of spreadsheet comprehension. As someone witnessing the power of metrics in other industries, I dedicated my career to becoming a Data-Driven Manager, and, as such, I began treating employees as data points, programs to be optimized, nodes on a workflow diagram. That naïveté I mentioned earlier reared its ugly head quickly and painfully as my team, while their productivity grew, began to lose morale, declaring that they were nothing but numbers in my eyes. And, honestly, they weren’t wrong.
My journey through management has been challenging and rewarding but, more than anything, it has proven to me how complex other humans are. I have learned countless lessons, and all were learned after a wealth of emotion, yelling, and sometimes even tears. As a mathematician, I know that things can always be better, a stylist can be faster, more precise in their recommendations, more perfect. As a manager who has had to testify in court for an employee to win custody of their child, or supported an employee through leaving an abusive ex, I know that reliance on data is simply not enough. Here are the major lessons I have learned in management and why Data can save us from bad mistakes but, without a touch of empathy, can easily drive us into new ones:
LESSONS FOR THE DATA-DRIVEN MANAGER
There are three numbers in management: ‘None’, ‘Like one’, and ‘Everybody’
There have been countless times that an employee has come to me stating, emphatically, “Everybody” wants this. People like to deal in absolutes. Generally, people might think, “If I feel this way, naturally everyone else does.” As members of the human race, it’s easy to assume everyone else is in the same bag. This means conversations in the office usually start with: ‘Everyone is saying…,’ because externalizing one’s needs and assuming everyone shares them is easier and less vulnerable than stating your own needs without modification.
Unfortunately, this assumed monolithic belief system loses what is essential to management: nuance. To treat every employee as though they have the same needs is one of the greatest pitfalls of a new manager. You know how to combat this? An anonymous survey, one of the cornerstones of data gathering from employees. Since our company began employing internal surveys we have been able to ensure that the loudest voice isn’t the only one that’s heard. Real life isn’t politics, we don’t stake our entire system of needs around a single theme. People have diverse opinions, some totally predictable and others utterly confounding. The power of survey is unparalleled when it comes to the emotional life of a company, as any good HR manager will tell you. Of course, as a manager, it can be hard to read unfettered and unfiltered feedback, but that’s reality. Which leads me to:
As a manager ‘no news is good news’ until everything is on fire
Managing a team can be an emotional experience. Not only do you have to deliver results, but typically we want our employees to be happy while they do so. That isn’t always possible, and yet, as data-driven managers it can be easy to focus only on the metrics that tell a good story rather than a true one. While ignoring nuance is a scary pitfall for any manager, confirmation bias can be a fatal one. I began my career staring only at sales numbers, lauding those stylists who could seemingly print money, and ignoring those who struggled. In the end, those disgruntled high-end stylists caused more damage to the company’s morale than the mid-range technicians who love their job ever could.
A manager’s job is to analyze reality and confront it, not avoid it. When used correctly, data is uncompromising and powerful. When used incorrectly, it can destroy companies and careers. While the real picture isn’t always pretty, it should be the basis of all decisions. When everything is good, it’s our duty to refute that truth until we can’t. A Data-Driven Manager should always ask, “What’s really going on?” which will usually answer the golden question: “Will things get better? Or worse?” By sitting back when things are going well, we’re never ready when the system breaks down. I have always managed to the belief that we should never trust a quick buck, because hunting down short term gains at the expense of long term profit is a fool’s game. If the present feels good, look to the past so you can ensure the safety of the future.
People are predictable, a person is not.
As a writer and a lover of stories, I suffer from a subset of confirmation bias I call Narrative Bias. There is a fascinating relationship between stories and human psychology and that can trick us into believing we’re living in an Oscar-winning screenplay. But narratives, while they can be predictive on a macro scale, can be extremely harmful on the micro level. Never assume that data controls the individual.
Early on, I believed I could predict people’s problems and solve them before they complained. I believed this was the mark of a good manager. What I didn’t realize then was that by removing the ‘problem’ before it was a ‘problem,’ I had stolen all agency from the staff member. People don’t like losing power, especially over something as private and personal as your own needs. As a mathematician, problem solving was usually about moving x to the other side of an equation and everything becomes clear. Try moving someone else’s stuff without asking, and you’ll realize real life isn’t so simple. There are a lot more variables, even if you’re only measuring one.
You can’t make everyone ‘happy’ but you can minimize annoyance.
This has been one of life’s hardest lessons. I’m a problem solver, so I have always assumed that if I solve people’s problems, they will be happy. QED. Well, that sentiment is sweet and horrendously wrong. Happiness is an emotion that can be unshakeable or as fickle as the breeze. It’s a fascinating and essential emotion we all need to feel in some aspect of our life. You know what happiness isn’t? A good business metric. In an industry as emotionally-driven as hair, it’s easy to link employee happiness with performance. We do it all the time. The issue is that when you manage to constantly increase everyone’s happiness, you can steer your business off a cliff.
Employee Happiness doesn’t care about profit margins until the company can’t pay rent. And don’t forget, not everyone needs the same thing. Trying to create policies to fit everyone’s individual happiness is next to impossible due to the fact that those needs can flip on a whim. Policies should not be magical, they should be fair. This counts for both internal and external policies. In the service industry, we have been chained to the adage, “The customer is always right.” Unfortunately, that phrase has been grossly misinterpreted. The customer BASE is always right. If 96% of your clientele wants something that is feasible…you should probably offer it. Unfortunately, on a micro scale, we find ourselves kowtowing to that leftover 4% because we lose all sense of the population when we’re confronted with a furious customer demanding something nobody else wants. Nobody wants to feel alone in their needs, and yet, no business should ever attempt to placate the entire population of the world. Data frees us to see the forest through the trees, even when the trees are yelling at you about customer service.
Everybody hates math
This statement is a facetious lie. On the contrary, everyone loves math, they just don’t know how to see it when it’s in front of their face. I learned very quickly that mathematics, in the end, isn’t a tool. It isn’t even a language. It’s a philosophy that has produced both of those things. Everyone ascribes to the philosophy of math, whether they admit it or not. On numerous occasions, I would sit down with a hair stylist and help them improve their business, we’d look at spreadsheets and equations, and they’d predictably respond with “I’m not a math person.” The numbers without analysis and communication are a spaghetti of nonsense. Ten minutes after these coachings, I’d see them walk to the desk and deftly navigate their books to maximize their dollars per hour, right before they designed a haircut by eyeballing angles and predicting cowlicks.
Math is a way of thinking, of quantifying, optimizing, problem-solving, and yet we get bogged down with the details of spreadsheet formulas. The best Data Scientists are communicators as well as analysts. It doesn’t matter how beautifully one can crunch numbers, if your team can’t interpret your work, you might as well have set your harddrive on fire.
SO, WHY DATA SCIENCE?
Data is essential to modern business. To say that you’re going to avoid data in your business is like saying you don’t need a smartphone in modern society. Sure, you can survive with just a flip phone, but you’ll always be at the disadvantage of speed and availability. I know, for myself, I don’t want to be left playing Snake on a Nokia 3310 when my competitors are all armed with the new iPhone.
I hated statistics in college for many reasons, but primarily because it didn’t give pretty answers. As a college student, that was infuriating. As an adult who has managed 100 individuals, I now realize: that’s life. Solving equations reveals truth in a perfect environment. Stats reveal truth in a chaotic one. Life is not perfect, as 2020 has plainly demonstrated, and the math we need to describe it should allow for that imperfection. Statistics will never be all-encompassing, but they can illuminate nuance we would have never expected. The world is complicated, we should reveal that complexity rather than simply shave away the outliers. As I said above, the greatest threat to any business manager is confirmation bias, and it’s a tendency of which we are all guilty. We like the answers we find comfortable. But comfort and truth are very different beasts.
I am embarking on a journey to become a Data Scientist because, as COVID-19 has ripped through every aspect of our lives and careers, I have discovered there is a whole new world waiting for us on the other side of this pandemic. The new world will have new metrics, new means, new medians, new playbooks, and we, as data-driven business managers, have a choice: settle for the pitfalls of conventional wisdom, or plumb the depths of data to create a new business environment. We don’t know what the future will look like yet, but given time, the right information, and a solid method, Data can make anything possible. Just don’t forget: everyone hates math. Data can’t do everything. If you lose the human element, you lose your team. And what’s a manager without a team?
