Developing a Field Guide to the Energy Efficiency Behavior Program Ecosystem

Blog Post | June 20, 2013 - 2:53 pm
By Susan Mazur-Stommen , Behavior program director

The Behavior and Human Dimensions staff at ACEEE have been conducting a form of ‘bird-watching’ by collecting sightings of utility-run behavior programs in the wild. We have been recording information about the variety of programs that exist, examining the frequency of their distribution, and generally seeking to ascertain how they come to occupy their specific ecological niches, including those at both electric and gas utilities.

Behavior programs seek to engage utility customers in efforts to change consumer decisions and actions, so that they may in turn save money and energy. To date we have collected over 300 different programs, run by more than 100 utilities during the past five years. In order to organize the diversity of offerings, we are using models borrowed from biology to form a classification system. One of the major goals of our latest project, which goes by the working title “Utility Behavior Landscape,” is the construction of a taxonomy of program types. This means assigning them to a hierarchy that is built upon principles that resemble those that give us family, genus, and species in biology.

The Utility Behavior Landscape is intended to provide a taxonomy (opposed to a typology) of behavior program types. One way to think about it is that typologies are analytic constructs, based upon ideal concepts (think Niels Bohr’s model of the atom that we studied in high school) whereas a taxonomy is generally based upon “empirically observable and measurable characteristics” with the analog here being a set of images taken of the atom using lasers.

There are three primary reasons for classifying behavior programs in this manner. First, we hypothesize that a mix, or stack, of program types is likely to deliver better results than will the reliance upon a single program type. Second, we believe that such a mix needs to be composed of program types that touch upon different facets of human behavior, including rational thought, emotional substance, and social interaction. Third, through this formal system of classification, we aim to show that program types differ in their effect and impact, which can help program implementers determine an optimal combination for their needs.

In addition to organizing and classifying behavior programs run by utilities, we are also examining their geographic distribution and areas of overlap. When we look at the ‘watersheds’ for each utility or third-party behavior-based interventions, an interesting picture emerges. While it may come as no surprise that some areas of the country are experiencing a ‘drought’ with respect to behavior programs, the idea that other areas are receiving over-watering may be new. There is a potential danger that message fatigue might set in for customers in regions with a high concentration of programs.

We further hypothesize that, when designing a behavior-based program, utility implementers should take the environmental context of their messaging into account. What other similar efforts may be underway in the same territory? Can a natural gas utility take advantage of groundwork laid by a campaign by an electric utility? Might it make sense to share data about the impact of certain communication efforts regionally?

We are still in the process of collecting and analyzing data, with more coming in every day. So if you have information about a utility-run, behavior-based program that you wish to contribute to this effort, please contact me at We would love to hear from you!