Political campaigns are armed with detailed knowledge about every prospective voter. Data-driven analytics tell them how to communicate with you.

Amazon’s product suggestions are often eerily on target. Its servers shadow you as you shop to learn about your interests and preferences.

Google knows you. Algorithms hang onto your every search term to determine which ads will appear as you browse the web and use your favorite mobile apps.

However you may feel about such avid admiration, marketers value it highly. Displaying an ad in your search results is one thing; displaying a perfectly targeted message at just the right time is another. Call it creepy if you will -- but marketers call it "market segmentation."

Technology makes it possible to keep dividing the market pie into smaller slices, until the treatment of every consumer becomes personalized. Intense data-crunching defines the planks of ever more sophisticated political campaigns, prioritizes cross-selling opportunities and otherwise leverages a growing mountain of data about every consumer.

But even before Google or Amazon, segmentation was a well-established practice in almost every organization, public and private. Consumer taste profiles guide new product development at Coca-Cola. Personas -- detailed descriptions of ideal but imaginary customers -- shape marketing messages at Levi Strauss.

Is market segmentation being used effectively to save energy?

“Segmentation is underused in energy efficiency,” according to efficiency researcher Jane Peters. Portland, Oregon-based Research Into Action, of which Peters is president, specializes in field research projects designed to better understand the motives and behaviors of energy consumers.

Energy efficiency is a key component in efforts to combat climate change. So why is an effective marketing technique being underutilized by energy efficiency marketers when so much is at stake?

The outreach around energy efficiency is largely driven by energy utilities. Even when campaigns are executed by a private firm or nonprofit, there are utility mandates and incentives behind them. 

Ratepayer relationships with utilities, from tariffs and billing to the type of meter on the wall, are based on rate classes. There’s the residential sector. Then there are small businesses and large commercial customers. And finally there are industrial customers. Efforts to engage consumers in energy efficiency initially stuck with those same segments.

“Utilities talk about using segmentation, but they’re usually basing that segmentation on rate class, and not on more meaningful attributes,” said Danielle Marquis, marketing director for SmartWatt Energy in Ballston Lake, New York. SmartWatt is an energy efficiency contractor for utility programs and large commercial, industrial and institutional clients. Its turnkey approach includes program marketing.

“Segmentation has been pretty much sector-specific, focused on residential, small business, large commercial and industrial, but no real work [has been done] to segment within those,” said Anne Dougherty, co-owner and co-founder of Illume Advising. Her firm, based in Madison, Wisconsin and Oakland, California, conducts research and designs energy efficiency strategies for contractors like SmartWatt.

“Although there’s been some movement, there’s not been a significant amount of investment in segmentation, because the industry hasn’t to date had trouble getting program participation, compared to the demands of the private sector,” Dougherty said.

Energy Trust of Oregon is a nonprofit that promotes energy efficiency and renewable power incentives to the customers of Portland General Electric, Pacific Power and other regional utilities.

“Our goal is to invest just enough to engage customers in our energy-savings programs, and so far, we have been able to do this effectively,” said Sarah Castor, Energy Trust’s senior project manager of evaluation. “As we have to go deeper to find energy efficiency resources, we will likely need to deploy additional tools, such as more sophisticated use of market research. “

Who is raising the segmentation bar?

A handful of innovative marketers are applying advanced segmentation in energy efficiency outreach, tailoring offerings for ever smaller segments of customers, and crafting messages to appeal to narrow categories -- and even to the individual consumer.

That’s not to say that utilities haven't been active in this domain. Large utilities and power suppliers have conducted attitudinal segmentation studies to better understand their customer bases. Bonneville Power Administration’s research in 2008 divided residential customers of utilities into eight segments. Working with Momentum Marketing Group in Cincinnati, Ohio, Bonneville leveraged an approach developed at Puget Sound Energy.

The eight segments describe the already-efficient “Green Idealists” at one end of the scale and the skeptical “Green Rejecters” at the other. Between those extremes are segments with names like “Disengaged Consumers” and “Comfort Is King.”

Unlike rate-class segmentation that focuses on buildings, attitudinal segmentation is about people. It helps marketers understand the segments they may have tapped out for energy-efficient upgrades and communicate to those that should be targeted next, based on a variety of demographics, psychographics and firmographics.

Targeting “segments of one” is challenging and expensive

It’s a big step from broad attitudinal segmentation to actual targeted marketing messages. For example, to deploy segmented marketing in a utility’s monthly statement stuffers, one piece about efficient lighting turns into a half dozen or more different pieces, inserted into selected envelopes depending on each customer’s profile.

Energy efficiency marketers would love to deploy the kinds of microtargeting and predictive modeling that Google and Amazon are using. They would like to have a profile for every prospect before they knock on the door, just like those political campaigners have. It’s often referred to as “the segment of one.” And that’s where segmentation gets really complicated.

In 2009, Energy Trust of Oregon researched customer segmentation, including psychographic research, and looked into four other segmentation schemes similar to Bonneville’s.

“We have not directly applied the marketing segments identified in that research,” Sarah Castor said. “Our marketing staff, for the most part, have been able to utilize messaging that appeals broadly to residential and business customers, and then target our offers to specific groups of customers based on energy usage, building characteristics, past participation with our programs, or other attributes.”

It’s easy to understand why marketers at nonprofits and smaller utilities don’t dive into using segmentation for targeting and messaging tactics. It’s technically complex -- and expensive.

“It’s mostly the program implementers who are doing more sophisticated work around segmentation,” said Anne Dougherty. “They often have performance-based contracts and incentives to realize savings.”

Austin, Texas-based CLEAResult designs, markets and implements energy efficiency programs on behalf of its utility clients. Laura Hutchings is the firm’s customer experience strategist.

“Segmentation usually is quite a large investment for a small or mid-sized utility,” Hutchings said. “But for an implementation firm, it makes a lot of sense, because we can make the investment and then scale it cost-effectively.”

A segmentation success story

Hutchings joined CLEAResult when it acquired her firm, Boulder, Colorado-based Populus, which had developed a customer engagement methodology for motivating people to invest in energy efficiency.

Populus was sending energy advisors into homes to propose and perform basic energy-efficiency upgrades. Their premise was that people’s attitudes mattered more than technical information when making investment decisions about their homes. The Populus tag line, “People first, buildings second,” was reflected in its selling style.

Almost all homeowners approached by Populus accepted the basic upgrades -- and Populus had a 70 percent success rate at persuading those homeowners to invest in more energy upgrades beyond the basics.

Populus asked its energy advisors how they were achieving such a phenomenal conversion rate.

“One of the things that we heard from energy advisors was that it was easy to frame messages and be persuasive when they really knew the customer,” Hutchings says. “By being in the house, they had all of these cues they could use to connect with the customer persuasively.” A Prius in the driveway, children’s toys in the living room, or a garden out back all provided knowledge that helped make the sale.

When Populus scaled up its energy advisor programs, replacing some in-home visits with phone and e-mail contacts, its challenge was to recreate that personalized approach without the lifestyle cues. To do that, Populus created propensity models by merging data about its past program participants with Experian data about homeowners. The models predict how much a given participant is likely to spend, and on what type of efficiency improvements.

When energy advisors look up a participant’s address, they see that person’s propensity score and know, with 62.5 percent certainty, which efficiency investments are the best ones to recommend. Like the campaign canvasser who knows how you’ll vote, CLEAResult advisors knows you’ll spend $1,200 for a more efficient water heater, before you even say a word.

Does the segmentation technique really work? “We’ve successfully transitioned our field-based model to a primarily office-based model without seeing a decline in conversion rates,” Hutchings said. The original in-home program, now conducted mostly remotely, still produces a 70 percent conversion rate. Those homeowners who convert make a $3,400 investment on average.

“For some of our propensity models, the likelihood of reaching the right customer with the right message is ten times higher than if we didn’t do any targeting,” Hutchings added.

Big data means a big commitment to IT

Big data analysis came into the limelight during the 2004 and 2008 elections, when Nate Silver’s FiveThirtyEight predicted voter choices with surprising accuracy. Not surprisingly, the more complex the segmentation, the more information technology is needed.

Opower is known in the utility industry for applying behavioral science to advance energy efficiency. The Arlington, Virginia-based company has invested heavily in the technical ability to slice and dice its database of 52 million households and businesses based on some 250 attributes. With its proprietary software, Opower can perform rapid segmentation on a utility’s customer list and target each segment with personalized tips by e-mail or text message the same day. The participation rate for those who receive targeted messages is 60 percent higher than for groups who receive blanket messages.

And the startup Faraday, which has built a microtargeting platform based on a machine learning algorithm, believes the same tactics used in political organizing can be applied to energy efficiency and solar.

Computing power has increased every year, while costs have fallen. More researchers can analyze more layers of data more efficiently than ever before.

“Some of the work Illume Advising is doing now for program implementers, we’re able to do at a lower cost than we could have four years ago because our software and computers can handle the data load much better,” Anne Dougherty said.

Danielle Marquis co-authored a paper earlier this year with Illume Advising founders Anne Dougherty and Sara Van de Grift, outlining recent experiments using segmentation to reach business customers. Illume developed propensity models for SmartWatt based on reams of data from more than 6,000 completed small business projects. The models are currently being used for market segmentation in SmartWatt Energy’s small business campaigns in four states.

To handle it all, SmartWatt developed its own software application, AMP, which enabled SmartWatt to organize all that data about program participants. The vast quantity of data available made the propensity modeling task possible. Today, salespeople know which customers to target with which program elements based on the propensity scores, which they see on their mobile devices.

“It would be next to impossible for utilities to build a propensity model like this, unless they’ve had a program running for a while,” Marquis said. “You need an extensive amount of past participant data, with a high level of granularity, to build these sophisticated models.”

Systems and data are essential, but having people who know how to handle big data can make or break a project.

“You can buy the computers, but if you don’t have the talent behind them, it can be a complete disaster,” Dougherty said.

People with those talents might want to dust off their résumés. Segmentation and microtargeting are gaining traction in energy efficiency programs nationwide.

“We are seeing more interest in adding these high-touch, high-tech engagement services as a way to lower acquisition costs,” Laura Hutchings said. “They can reduce the amount they spend on marketing.”

Even the attitudinal segmentation that has been done around energy efficiency will need to be kept up to date.

“The problem with segmentation is that it’s not evergreen,” said Dave Moody, marketing lead at Bonneville Power Administration. “If you want to get the most value out of your segmentation, given the mobility of people and changing economic drivers, you have to make sure your data is up to date.”

Looking forward

“The future of segmentation is ever smaller segments and more personalization,” said Emily Bailard, Opower’s director of energy efficiency solutions marketing. She sees the leading edge of segmentation increasingly resembling consumer marketing practices outside the utility industry. Opower’s latest software platform can target customers at life’s turning points with relevant offers described according to each customer’s attitudes toward saving energy.

“Amazon’s recommendation engine is something we aspire to and that we hear utilities talk about,” Bailard said. “But it’s not just about delivering the right offer. It’s critical to make sure the recommendations come at the right time for that person.”

Marketers in every industry wish they knew as much about their prospects as they know about their existing customers. For utilities, every prospect is an existing customer. Energy-efficiency outreach programs are gradually starting to treat those customers less like buildings and more like people.

“Our approach has been to turn things upside down by putting people first and the building second,” said Hutchings, “then framing how we engage a customer based on that -- thinking about storytelling rather than energy audit reports. It’s thinking differently about how we connect with customers.”


Denis Du Bois writes about sustainable energy and advises energy technology companies on marketing as a partner at P5 Group. He lives in Seattle and has an off-grid solar getaway in the Cascade Range.