The calorie label on your Coke is a failure of behavioral design. Here's what the energy industry can learn from it.


banner 1.2


Opower's company roots are pretty humble: we started as a conversation about kilowatt-hours. Specifically, we wanted to motivate people to use fewer of them. And not with incentives or devices, but instead with information — the basic facts about your energy use over time, contextualized in a way that motivates you to use less.

The hitch: the standard unit of measure in the energy industry doesn’t mean much to most of us. What is a kilowatt-hour, really? We knew early on that we’d need a proxy — something that ordinary consumers both know well and care deeply about. That’s why our Home Energy Reports contextualize energy consumption in terms that matter. Light bulbs. Dollars. Your peers. So our ears perked up when we heard public health researchers striking up a similar conversation this month. The American Journal of Public Health just published a new study exploring how calorie labels, which offer information intended to influence our eating habits, actually affect behavior — and how we as a society can design effective ways to promote healthy choices. Here’s how Vox's Sarah Kliff explained the experiment:

“Researchers at Johns Hopkins University wanted to see what would happen if they made calorie labels more blunt — specifically, if the labels told consumers how much exercise it would take to burn off the energy in a soda.

They went into Baltimore and put up signs in six corner stores' fridges, right where the drinks were. They said it would take 50 minutes of running or 5 miles of walking to burn off the 250 calories in a 16-ounce soda. And they waited, over six weeks, to see if shoppers made different decisions.”

Lo and behold, they did. And the results were actually pretty remarkable. People in the test group were far less likely to purchase soda or sports drinks, and far more likely to choose water — or no drink at all — relative to a control group.


Researchers measured how different messaging about calories affected Baltimore corner store shoppers' purchases over a six-week period. People who received an exercise-related message ("intervention," light gray) were less likely to buy soda and sports drinks than people who did not ("baseline," dark gray).

The finding is particularly relevant now, as restaurants nationwide start posting calorie labels next to their entrees. If we really want to get the most bang for our buck with labeling, the research suggests, we’d be better off telling people how long they’d spend working their burger off at the gym — not how much energy is in it.

"To really motivate behavior change, you need to frame consumption in a way that's both universal, so anyone can understand it, and personal, so it actually means something to real people."

That finding aligns well with our own research. To really motivate behavior change — which, for us, means getting homes and businesses to use energy in ways that are efficient, affordable, and well-timed — you need to frame consumption in a way that’s both universal, so anyone can understand it, and personal, so it actually means something real to people. So we don’t talk about kilowatt-hours in isolation. Instead, we use data analytics to help utility customers understand how long it takes for an efficient light bulb to pay for itself; whether they used more natural gas this October or last; if their neighbors down the street are saving more electricity than they are. Things that matter, and motivate us. That’s how we’ve helped more than 50 million people save more than 5 billion kilowatt-hours of energy. And it’s why we don’t let abstract concepts like that stand on their own. We’ll tell you that 5 billion kilowatt-hours is enough to take nearly a million cars off the road, or power New Hampshire for an entire year.



Feature image credit: Flickr user namoscato.

Be the first to comment

Comments ( 0 )
Please enter your name.Please provide a valid email address.Please enter a comment.CAPTCHA challenge response provided was incorrect. Please try again.