:format(webp)/cdn.vox-cdn.com/uploads/chorus_image/image/63708420/fyd_detox_tipp_data-diet-470x329.0.1485548634.0.jpg)
Maybe you’ve heard of — or even tried — the Atkins Diet or the Paleo Diet or the Mediterranean Diet. Or the South Beach, Ornish, Asian or Flat-Belly Diet, or any of the hundreds of others that come and go and come again, as fads.
But I’m here with a new approach: Let’s call it the Data Diet.
Don’t laugh. I lost 40 pounds on the Data Diet. And then when, like so many, I gained it back over time, I lost 30 pounds again late last year over roughly the same amount of time.
I am a data guy, an engineer by training. I’m knee-deep in data every day. I co-founded a company that uses big data, machine learning and natural-language processing to help make the Web more relevant for consumers and for those who do business on it.
I am a data guy, an engineer by training. So maybe it’s not a shocker that I turned to data nine years ago, when I discovered that my cholesterol level was slightly elevated.
So maybe it’s not a shocker that I turned to data nine years ago, when I discovered that my cholesterol level was slightly elevated. I got the news at a company health fair at Google, where I was working at the time. I was told that I was on a course that would eventually require cholesterol-lowering medication. That wasn’t for me.
After doing some research, I reasoned that weight loss was the most reasonable alternative to drugs. I upped my exercise routine. I became more cautious about what I ate, and I ate less of it. (Hey, garbage in; garbage out.) And most importantly, I bought a scale and weighed myself — every day.
In engineering, simplicity is often seen as a virtue, and admittedly my approach was fairly simple. But, if I may say, that’s part of what makes it effective. Essentially, I established a measurable metric that had a direct correlation to my goal, and determined the proper interval at which to measure it.
And as I carried out my weight-loss campaign (which I’ll get to in a minute), I realized that the lessons learned using data in life apply equally to using data to run a business. Specifically:
- Don’t overcomplicate your reliance on data.
- Be clear about what you want to measure, and understand the steps you’ll take depending on what the data tells you.
- Don’t allow meaningless data points to become a distraction.
- Establish an appropriate interval for measuring the progress, or lack of it, that you’re making on the way to your goal.
- Use common sense. Data is powerful, but completely disregarding human intuition and experience can diminish its value.
So, how did any of that help me lose 40 pounds? Every morning, I’d step on the scale. If my weight had decreased, I kept doing what I had done the day before. If it stayed the same or increased, I’d either eat less or exercise more. And in time, I lost a significant amount of weight.
Yes, I had tried to shape up before with diet and exercise, but with disappointing results. The difference this time was that I had a repeatable and meaningful way to chart my progress. My daily appointment with the scale eliminated any doubt about whether I’d stuck to my program or not.
It’s no wonder that Fitbit and wellness apps are all the rage. Data is a powerful motivator. We live in a measurement society.
Even for me, a data guy, it was a bit of an epiphany. It’s no wonder that Fitbit and wellness apps are all the rage. Data is a powerful motivator. We live in a measurement society. Everyone talks about metrics. Businesses live by them — P&L, ROI, gross margin, churn, customer lifetime value. Our jobs sometimes depend on hitting a number. We consider school-test scores when we decide where to live. Parents talk about their kids’ GPA and SAT scores.
So why not focus on metrics as a way to affect positive health changes?
I said my system was simple, which isn’t necessarily the same as easy. First, I’m sorry to report, losing weight still takes willpower. It’s one thing to know what you must do. It’s another to do it.
Secondly, turning to metrics is never the whole answer. Whatever you’re setting out to do, you need to select meaningful metrics and devise the right frequency for analyzing them. Numbers don’t lie, but they can be manipulated, misconstrued or even meaningless.
Think about it: Say I had chosen to weigh myself every 15 minutes. Aside from the disruption to my day, that sort of frequency would lead to false assumptions. What if, for instance, in 15 minutes, I had done some vigorous exercise while hydrating with a quart of water? Net result: A one-pound weight gain. Should I stop exercising or stop drinking water? Should I stop both?
Of course not.
Or what if I weighed myself once a month? How practical would it be to recreate an entire month of meals and activities in the event my monthly weigh-in yielded positive results. Conversely, if I gained weight, how could I cut through the noise generated over a month of living to determine what to scale back and what to ramp up?
The same is true in business, of course. Parts of your business probably move too fast to measure certain metrics only once a month. But measuring most metrics hourly provides noise and a fuzzy picture of what’s going on fundamentally.
And maybe best of all for a data guy, I got a repeatable result that proves the efficacy of my approach.
Why am I telling you all this now? History has repeated itself. After my initial 40-pound loss, I kept the weight off for years by weighing in every day. If my weight was up, I cut back on food. If my weight was lower or the same, I relaxed as far as diet restrictions.
But then life happened: I got married, co-founded a company, had a child. My system unraveled, and I gained back 30 of the pounds I had lost. Earlier this year, during a routine physical, I learned that my glucose level was slightly elevated. No cause for immediate concern, but it bothered me, because I know that South Asians have a higher incidence of diabetes than the population as a whole.
It was time to return to the Data Diet — and to cut out added sugar altogether. Again, I started weighing myself every day, and started seeing my weight drop. It’s been months since I returned to my target weight. My blood sugar is improved.
And maybe best of all for a data guy, I got a repeatable result that proves the efficacy of my approach.
Ashutosh Garg is a co-founder and board member at cloud marketing platform company BloomReach, and a true guru of all things search, with 10 years of experience in information retrieval, machine learning and search. Previously, he was a staff scientist at Google, and at IBM Research prior to that. He is also a prolific publisher/inventor, with a book on machine learning, more than 30 papers, and more than 50 patents. Reach him @bloomreachinc.
This article originally appeared on Recode.net.
