The expression “keeping up with the Joneses” usually means consuming more of something. Through an innovative effort, Connexus Energy has tapped into its customers' desire to emulate their neighbors so that they consume less electricity. In doing so, customers have saved money on their electric bills, and Connexus Energy has benefitted by meeting a state goal to reduce power use — and in a cost-effective manner.

New Challenge

Connexus Energy is the largest member-owned cooperative in the state of Minnesota, providing electric service to 125,000 customers in seven counties near Minneapolis-St. Paul. The utility has long offered a host of conservation improvement programs (CIP) to meet goals, including high-efficiency appliance rebates, compact fluorescent and light-emitting diode (LED) lighting rebates and air conditioner cycling programs.

Connexus was presented with a new challenge in 2007 when the Minnesota legislature passed the Next Generation Energy Act requiring all utilities operating in the state to reduce their energy sales by 1.5% per year. Although the requirement was not set to take effect until January 2010, Connexus officials realized early on that traditional demand-side-management (DSM) programs were only going to go so far in meeting the 1.5% target. Clearly, a new approach was needed, so the officials started working with the Minnesota Office of Energy Security (OES) to develop some ideas.

For assistance, OES staff brought in OPOWER, an energy-efficiency software company. During the meetings, Connexus learned about OPOWER's Home Energy Report program, which combines advanced analytics with behavioral science techniques to drive energy savings by providing customers graphs on household energy use by month, quarter and year, and then comparing energy use against peers.

A Novel Approach

This novel approach is based on work by Robert Cialdini, OPOWER's chief scientist and a professor and social psychologist at Arizona State University who has long studied what motivates people to take environmentally responsible action. Cialdini's studies indicate that facts alone often do not prompt people to change their habits and save energy. Instead, his studies show that consumers are more likely to respond to normative messaging — motivating consumers by letting them know what others are doing through meaningful messages that get their attention and spur them to action.

For example, telling customers something like “you used 10% more electricity than your efficient neighbors” or “most people in your area keep their air conditioning at 78°F” is more likely to change their behavior than simply asking them to save energy for the environment. It also helps to give consumers customized information on how they can reduce energy use.

Today, OPOWER works with about four dozen utilities around the United States. But at the time Connexus contacted them, only two other utilities — Sacramento Municipal Utility District and Puget Sound Energy — were using Home Energy Reports. Thus, it was a little scary to get on board. However, Connexus has always considered itself innovative, and the possibility of achieving significant energy savings at a relatively low cost to the utility (compared with traditional DSM programs) was tempting. Connexus signed on to conduct a 12-month pilot program with its residential customers — becoming only the third utility in the country and the first in the Midwest — to use Home Energy Reports.

The Pilot

The pilot began in March 2009 with 40,000 customers in a test group and another 40,000 in a control group. Home Energy Reports were mailed monthly to some customers in the test group and quarterly to the others. The control group received no reports, but their energy use was monitored.

The Home Energy Reports, which were mailed separately from bills and made available online with a secure login, rated customers' energy use against nearby neighbors in approximately 100 homes of similar size that used the same heating fuel. The reports compared a household's kilowatt-hour use in the previous month or quarter with “all neighbors” (the nearby 100) as well as “efficient neighbors” (the top 20% energy savers among all neighbors).

Those on par or doing better than all neighbors saw a smiley face on their bills, while those on par or performing better than efficient neighbors saw two smiley faces. Along with the monthly or quarterly comparisons, the reports also charted kilowatt-hour use over the previous 12 months versus all neighbors and efficient neighbors — citing (as a percentage) more or less electricity use and extra cost or savings, as applicable.

Equally important, the reports provided personalized action steps to reduce electricity consumption based on a customer's past use and housing profile. Typical tips featured smart purchases, such as buying efficient lightbulbs and occupancy sensors, and quick fixes, such as turning off lights when not needed and running appliances off peak.

Advanced Analytics

Charts and statistics in the Connexus Home Energy Reports reflected analysis of customers' energy-use patterns. OPOWER looked at monthly billing data, program-specific characteristics, customer demographics and climatic data. In turn, Connexus Energy provided data to Power System Engineering Inc. (PSE), an independent third-party evaluator hired to validate the methods and estimates of energy savings attributed to the pilot program.

PSE calculated energy savings in three ways:

• The true impact test, which examined the change in the differences of the control and test groups from pre-pilot to post-pilot time periods

• An ordinary least squares (OLS) econometric model

• A fixed effects econometric model.

All three evaluation techniques are standard methods of measuring energy-efficiency impacts and all three showed similar results, with the true impact test estimating a daily kilowatt-hour reduction of 0.621 for each member of the test group during the post-pilot time period. Similarly, reductions of 0.622 and 0.637 were seen for the OLS and fixed effects models, respectively. Reductions ranged from 2.05% to 2.1%, with an average of 2.07%.