Usage-Based Segmentation Can Create Value in Electricity Markets

Every New Year brings pronouncements about energy trends, and for the past few years, “the year of the customer” has repeatedly been declared. In 2016, that assertion came from Utility Dive, which highlighted the expansion of behavioral demand response, electric vehicles (EVs), and third-party energy services for homes and businesses as key trends pointing to the increasing centrality of customers in the electricity system.

Major market participants, including utilities and distributed energy resource (DER) providers and their regulators, have been trying to better understand and leverage data-rich knowledge about customers in recent years, echoing similar efforts in retail shopping and other industries. More than 50 percent of utilities, including new entrants like TFC Utilities and existing players like ComEd and Duke, report moving to a more customer-centric business model. New regulatory regimes are arising in California, New York, and elsewhere to focus industry efforts on customers, and DER providers are creating new units that require empowered customers for success.

Central to these customer-engagement efforts is the need to better understand who customers are, what they care about, and how their priorities might be met through new DER products and services or different marketing outreach. To achieve this understanding, utilities, DER providers, regulators, and other industry stakeholders will need to forsake traditional sector-based labels such as “low-income residential consumer” and “large commercial consumer”—and, increasingly, basic demographic and attitudinal segmentation—and adopt advanced usage-based segmentation. Usage-based segmentation divides customers into groups based on their actual patterns of energy use, helping DER providers and utilities to do a better job of meeting actual, observed customer and grid needs.

Segmentation: An Important Tool to Engage Customers

To move toward a more customer-centric business model, companies need to have a basic understanding of who customers are—and that’s where segmentation analysis comes in. In the DER context, segmenting means separating the diverse population of end-use individuals into subgroups with similar needs and preferences. Those needs and preferences can be identified through surveys of demographic or attitudinal characteristics, but they can also be measured through energy-usage behavior, where such data is available.

Where patterns emerge, energy market participants can better tailor products and internal strategy to meet those customers’ needs, increasing the overall adoption of a targeted DER product or service and increasing the cost-effectiveness of marketing and outreach efforts to encourage that adoption. This is at the heart of most marketing strategies across industries—think of the targeted ads on Facebook, or the differences in the products and layout of stores in urban cores versus suburban shopping malls.

Sectors Are Not Segments

In the past, sector- or income-based customer labels were used for discrete ratemaking and rate-allocation, ensuring relative equity and rate stability and avoiding rate discrimination. They worked well when customers appeared to have similar economic concerns and responsiveness to rates. However, sector-based labels no longer accurately serve increasingly diverse customer classes. Take the example of low-income residential customers, a sector that includes large families, elderly homebound residents, and individuals with home-based businesses, all of whom could have drastically different patterns of energy use and adoption of distributed energy resources and new rate structures.

Some utilities have used more complex tools to divide customers into attitudinal or demographic segments, with some going so far as to segment based on similarities in customer decision-making frameworks (in what’s called psychographic, or lifestyle segmentation). For example, research out of California suggests that there is a high correlation between EV drivers and rooftop solar owners, so people who purchase rooftop solar may also purchase EVs. Many DER providers use these advanced tools where possible, although in many markets, they have clamored for increased access to customer usage data in order to better understand actual customer needs. In contrast, with the exception of leaders like PG&E, utilities have not employed usage-based segmentation, that is, labeling of customer groups based on actual data about energy-use patterns.

Like market participants, regulators rarely make use of advanced segmentation for rate or program design. New York Public Service Commission Chair Audrey Zibelman recently said:

I think we need to become much more discrete about the kinds of customers we’re talking about, who are more price conscious, more environmentally conscious, younger demographics, customers who want simplicity, all those types of things.…Regulators are really lousy at marketing. We really don’t understand customers; we think about things in big flavors—there’s vanilla, chocolate, and strawberry, and that’s about as exciting as we get.

Without actual data about energy use, and with attitudinal and demographic data constrained by the cost of surveys, utilities, regulators, and DER providers have had limited knowledge about whether particular customers would benefit from different DERs, rates, and types of engagement. However, the increased availability of energy-consumption data through advanced metering infrastructure (AMI) can enable real observation of use through non-intrusive load monitoring and enable greater responsiveness to the changing needs of customers.

The Changing Utility Customer

Three trends suggest the need for labels that take into account more individuated grid and customer needs in order to enable a different approach to customer engagement. Utilities, regulators, and DER providers should choose levels of granularity that vary based on the purpose of the segmentation, as well as on the cost of segmentation itself.

1. With the availability of more granular information, researchers better understand how diverse and individuated customers are, noting that current models have poor predictive capabilities that miss important targeting opportunities. The growth of telecommuting, home-based businesses, and multigenerational households, combined with increasing customer adoption of distributed resources such as smart thermostats, rooftop solar, and grid-interactive water heaters suggest that energy-use patterns are evolving. As a result, current rates, programs, and marketing strategies may not reflect customer realities, decreasing the likelihood that customers respond to existing DER product marketing and price signals in rates.
2. Customers expect two-way communication from their service providers. Companies like Amazon and Netflix use customer’s web browsing information to provide customized recommendations that anticipate needs. Even organizations in more traditional industries, such as banking, the postal service, and taxicab services, have created applications to enable interactions with customers. Given the ubiquity of customer-provider communications, customers are more easily frustrated by antiquated systems that make irrelevant or poorly timed suggestions or offers. For DER providers and utilities, communication with customers will be meaningful and provide value only if it references real customer pain points.
3. Customers are increasingly comfortable with automation and value-based pricing. When customers book an airplane or concert ticket online, they expect higher prices during times of higher demand. When DER providers and utilities employ usage behavior to segment customers, they can use this comfort with value-based pricing and automated devices to create value for those customers and the grid. Regulators and utilities can enhance customer responsiveness to interventions like more granular time-, value-, or location-based pricing by targeting the customers most likely to benefit from and respond to those programs.

New methods of customer labeling can leverage these customer trends to enable DER as a core grid resource by improving the effectiveness of rate design and DER program and product marketing.

Discrete Segmentation Enables System Cost Savings Through Rate Design

If segmentation is more discrete and tied to usage, utilities and DER providers can better serve increasingly differentiated energy-services needs by proposing rates, services, and products that fit actual behavior. Social science research suggests that current energy usage is one of the best predictors of energy savings from behavioral programs. In other words, actual consumption information may be more predictive than other attributes of the value of DER services and rates. For instance, in order to target demand-response and energy-efficiency programs to certain customer segments, Stanford and PG&E used AMI data to identify customer segments with similar load profiles, and then layered attitudinal and demographic variables on the segments.

Texas utilities could use a similar approach to customer segmentation. With low nighttime wholesale prices and high peak summer prices, Texas utilities have begun offering a variety of creative rates, such as TXU Energy’s Free Nights plan, designed to shift energy consumption to times when wholesale prices are lowest. These utilities could use targeting based on load profile to identify those customers most likely to adopt the new rates, creating marketing value for their competitive retail business and reducing system costs.

Timely Segmentation Facilitates Better DER Targeting

If segmentation is done in a more timely fashion, it can increase the ability of utilities and DER providers to respond to changes in customer’s lives. Usage changes may indicate recent life changes, such as a new baby, a child moving away to college, or a new home-based business. DER providers and utilities armed with timely data about such changes can create opportunities to better serve customer needs by readjusting rates, targeting behavioral efficiency interventions that reduce customer costs, or promoting DERs that better serve new needs. However, privacy concerns may create challenges for utilities and DER providers attempting to use this type of customer information to deliver customized offerings.

More broadly, rate design itself could respond much more rapidly to trends within a service territory. The current process for rate design typically involves slow, inflexible regulatory review of infrequent rate cases, which results in approval for one rate with little or no provision for that rate to evolve over time. Regulators could create processes that enable more rapid responsiveness to the need for new rates, in which pilots and experiments that create value for consumers and the grid have clear pathways to become widely available options.

Shared Data Enables a Broader Range of Innovation in DER Marketing

Although utilities have a role in enabling and perhaps performing better segmentation, other market entities may be better positioned to extract value from more discrete, timely data. If the data that drives segmentation is more flexible and is shared in platforms, a wider range of market participants can use it to market to and segment customers. Over the past few years, the market focused on energy disaggregation using AMI data has grown rapidly, with a variety of business models that target utilities, DER providers, contractors, housing authorities, and others. Some utilities have created application programming interfaces (APIs) or “one-click” tools, such as the Green Button Connect My Data Program, to enable customers to share data, but access to consumption data remains limited in many states. A recent Mission: Data report highlighted this gap, noting that “many innovative companies spend a great deal of time and money accommodating each utility’s idiosyncratic differences with regard to data access; even worse, many companies are forced to install their own electric meter on the customer’s premises, redundant to the utility’s, in order to get usage information in the manner they require.”

The availability of this data is an emerging area of interest in DER reform debates in multiple states. In California, regulators have been addressing data availability since 2010, with interest in the topic continuing in response to the utilities’ recent distribution resource plans. A New York Reforming the Energy Vision proceeding is also addressing the availability of data in shared platforms, asking for input on customer- and aggregated data provision at a December workshop, and providing utilities with an opportunity to rethink customer-data sharing as a part of the distributed system implementation plans guidance. Value-added data analysis may be one new type of alternative revenue stream for utilities moving toward more market-based DER procurement.

Looking Forward

Although usage-based segmentation offers the opportunity to enable more rapid, grid-supportive DER adoption if it is more discrete, timely, and available, market participants will have to grapple with privacy and cost-effectiveness concerns. Current privacy rules vary by state and tend to restrict the easy flow of customer-usage data to market participants. Similarly, protocols for accessing, storing, and using such data vary, which can affect whether the value of segmentation exceeds the cost.

Proceedings and workshops in states with comprehensive DER reform efforts offer stakeholders and market participants the opportunity to engage with these questions and open up segmentation as a tool to enable customer DER adoption and grid stability and resilience. The increasing availability of granular energy data and improved segmentation tools creates the opportunity for utilities and DER providers to initiate experiments to better understand what customers really need and how they will behave, enabling value creation for individual end users and the grid as a whole.