Behavior Change
Follow-Up to SB'13:
Lucid's Michael Murray on Using IT to Drive Behavior Change

Michael Murray is the president and co-founder of Lucid, a software company providing real-time information feedback designed to “teach, inspire behavior change, and save energy and water resources in buildings.” Since the company’s founding in 2004, over 275 schools and organizations have adopted Lucid’s Building Dashboard, which it describes as a social network for buildings that allows people to view, compare, and share resource usage data online.

Murray presented some interesting lessons from his work during a panel discussion at Sustainable Brands ’13 last month in San Diego, called “Data and Behavior Change: Is Seeing Believing, and Is That Enough?” I followed up with him to kick off Sustainable Brands' new IT and Big Data content channel.

Can you give an overview of how Lucid’s software is being used in the marketplace and a rough estimation of how many people are engaged with it?

Lucid’s Building Dashboard started in green buildings and in universities. We have all of the Ivy League schools and many LEED-certified and “net zero energy” buildings using our software. But it has also spread to state and local governments, who have to manage a lot of facilities and large utility bills, as well as corporations such as Google and PNC Bank. Earlier this year, we ran the third “Campus Conservation Nationals” with the U.S. Green Building Council. This is a fun competition to see which university can save the most energy during a focused period of time. We had over 200 universities take part, with several thousand buildings and over a quarter of a million students participating.

Research shows that simply providing people with feedback about their resource consumption can lower consumption rates. But what is required in order to teach or inspire them to go even further?

Human psychology is fascinating. There’s a disconnect between conscious thoughts and attitudes on one hand, and behavior on the other. A lot of behaviors — some would venture to say most — are unconscious. So one question is: If we can successfully spark conservation behavior, does that change people’s minds in any meaningful way?

Through our research partners at Oberlin College, we have some evidence that we can increase people’s “connectedness to nature,” which is correlated with pro-environmental behaviors. But first of all, we must ask ourselves if changing conscious attitudes is really a worthwhile goal. Do we care about pro-environment sentiment if we can successfully nudge people to make the right decisions in their day-to-day lives? After all, it is possible to do the right thing for the wrong reasons (e.g., one can buy green energy because it was marketed very aggressively, and not because you genuinely are passionate about the environment or climate change). I think the jury is still out on that one.

How important is data visualization in your work, and are best practices well-established? Or, are you still experimenting with ways to better communicate data to users?

Electricity is often referred to as “the invisible commodity,” so we have to constantly experiment with new ways to communicate quantitative information. Data can be very persuasive when presented in the right way. Our best-selling product, Building Dashboard, has won numerous design awards, and all of that credit goes to our co-founder and Chief Creative Officer, Gavin Platt. A few years ago, we moved beyond simple data presentation to socially and environmentally contextualized information. What that means is presenting one’s consumption in the context of your peers, or in the context of environmental impact. For example, showing your building’s energy consumption ranked against others in your organization.

Not surprisingly, social norms exert a very powerful force on behavior. The next visualization challenge for us is demystifying some of the very technical energy analysis tools used by facility engineers to manage buildings and HVAC equipment. Making those tools useful and accessible to non-technical people means you can tap into the initiative and desire to act from a broader group of individuals. Energy management is more successful when everyone is an energy manager, not just a single person in the basement who does maintenance and doesn’t really have time to be proactive.

At SB’13, you said the resolution of the data provided matters when incentivizing participants in energy-reduction programs. What do you mean by that?

In our first peer-reviewed study, we found that buildings metered at sub-one-minute intervals saved a lot more electricity (55%) than those with weekly meter readings (about 30%). What that means is low-resolution data can still make a difference in the right environment, but real-time data takes it even further. For example, it’s really empowering to be able to turn off lights, computers or appliances, and almost instantly see the reduction displayed on Building Dashboard. It makes people feel like they can actually accomplish something. In contrast, it can be hard to make people feel like recycling plastic bottles really makes a difference when you see those pictures of landfills and the plastic circling in the ocean; it feels like just a drop in the bucket.

When it comes to providing high-resolution feedback, are there significant data management or analysis hurdles to overcome? My understanding is that the field of data analysis is still relatively immature compared to our ability to collect data, but I’m not sure what that means in practice.

Actually, I think the opposite is true. Nobody likes to talk about the data acquisition problems, but it’s really the Achilles heel of the industry. That’s why we often say that “Big Data” in commercial energy efficiency is like teenage sex: everyone is talking about it, but very few actors in the market are actually doing it.

The data acquisition and management challenges in energy management are enormous. On the face of it, it doesn’t seem that hard to get time-stamped kilowatt-hour readings into a database. There’s software that does that, right? But the reality is much different. There are myriad equipment manufacturers, each with different protocols and communication methods. You can’t simply plug an electric meter into your local ethernet cable and call it done. There are gaps in data, bogus readings, network intermittencies and all kinds of pesky problems that stand in the way of getting consistency and reliability. You could have the greatest analysis tools in the world, but if your data record is poor, it’s all worthless.

In San Diego, you said data was important, but not primary, when engaging users. What is primary?

Social norms are the primary engagement tool. If your peers are doing a hundred things and getting LEED and Energy Star on their buildings, and you’re not, then you start to look like a chump. Where the data comes in is in linking cause and effect. Did that action actually reduce energy use or not? Time and money are scarce, so you need a way to judge your own effectiveness. Another way to think about it is in the context of health and weight loss: You can get a scale in your bathroom, but you won’t use it as a tool unless you’ve first decided to do something about it.

Have you found that people take naturally to feedback systems, or do they have to be well-marketed?

Yes, they have to be well-marketed. With the oceans of information on the web today, having a good portal doesn’t mean that people will use it. That’s why we’ve integrated pieces of Building Dashboard into our customers' well-trafficked websites in order to draw attention. We also create fun contests because people love competing with each other. The point, of course, is not the competition itself, but the attention and deeper thinking about energy that results.

Isn’t there some inherent demand for feedback in the user experience, too?

For the average consumer, there is an initial interest in using feedback technology. Think of the Prius’ mpg gauge. But then the usage tapers off as interest wanes. That doesn’t mean the feedback is no longer effective; it could mean people’s habits have been formed, and they don’t need to rely on the feedback as often. Of course, we also hear about the benefit of device automation in the long term, because people don’t have the interest to actively manage energy usage of every device in their lives. Letting control kick in at the right time is perfectly sensible.

What’s different about an organization — a company, a school, a state government — as opposed to a consumer is that the individual who makes resource-use decisions in an organization has a boss, and that boss cares about the utility bills. So the first benefit is the Hawthorne Effect, which is the observation that systems behave differently when they’re being watched, even if there isn’t any direct manipulation. If you know you could be called out by your boss for poor performance, you’re probably going to manage things better, even if it’s unconscious.

The second benefit is a cut-and-dried accountability tool. A lot of our feedback tools are really just quantitative measures of job performance. The facility manager has to present monthly or quarterly numbers and our technology is what generates those reports. There is a very immature business process around energy management today, so good reporting tools are the first step.