Context
On most weeknights, most people are returning home from work or school between the hours of 4pm and 8pm. With so many people home and using electricity, energy demand spikes during these “peak hours” and puts a strain on the energy grid. Utilities would typically have to resort to “back-up power plants” to meet energy demand, which are costly to operate and worse in terms of greenhouse gas emissions. Utilities in the US also have to comply with regulations requiring them to reduce energy use during these peak hours, so many began switching to “time-of-use” rates — where energy has different costs through the day to encourage lower use at this time — instead of traditional flat-rate or tiered pricing.
This scene from the 2008 documentary Britain From Above illustrates the peak energy demand conundrum in the most British way possible: watch the clip ↗
Business needs
Many utility partners weren’t seeing the drop in peak energy demand they were hoping for, and any communications or education about time-of-use rates wasn’t landing with their customers.
User needs
Most utility customers didn’t seem to have a full grip on their peak hours, and even fewer understood how time-of-use rates even worked. What could actually be more beneficial for them wasn’t resonating with the communications their utility was putting out.
The operative question became: how do we reduce energy demand at scale using existing communication channels?
Approach
Utility customers in the US spend about 9 minutes per year thinking about their utility, according to Accenture. And while that sounds brutal, it’s actually table stakes for this and every report we built at Opower.
Our initial approach was to insert peak-hour messaging into existing reports, like the High Bill Alert, which triggers when a customer is trending toward a higher-than-normal bill. The version we tested paired usage data with a message like this:
You’re on a time-of-use rate plan that charges more for energy used from 5pm to 8pm. By using less energy at this time, you can save more on your bill.
For users who didn’t know what a time-of-use rate plan was, this message felt poorly timed and out of context. We spoke with utility customers whose utility had recently switched to time-of-use rates and found that many didn’t know they’d switched — and even fewer knew how a time-of-use rate worked.
In order to move the needle, we needed to build a more bespoke report designed specifically for time-of-use customers and the utilities that needed to reduce peak energy demand.
My work
As the content lead, I worked closely with the UX Designer and Product Manager to develop the report experience. We decided on a weekly email series that would give customers the opportunity to make changes during peak hours and see how it affected their energy use week to week.
The content strategy largely revolved around this cadence, starting first with an intro email that would introduce the program and lay down the foundation for time-of-use rate education.
In the headline, we introduce the email as their “Rate Coach” to set the aspirational tone, with a short description that highlights both the primary benefit (saving money) and how they’d achieve it (avoiding high-cost times). This headline plants the seed that “high-cost times” exist, using loss aversion to sell the value of the email.
The subhead took the most amount of time to get right. This subhead, along with the header, needed to deliver the most critical sequence of information to lead to the ultimate goal: get customers to use less energy during peak hours.
To determine the right sequence, I initially took a maximalist approach and included all of the information and context that might possibly be relevant to the user:
Your utility has changed to a new rate plan called a time-of-use rate to reduce the need for costly back-up power plants during the times when energy demand is highest. These “peak hours” occur when a lot of people are all using more energy at the same time, and happen every weekday from 4pm to 8pm. During peak hours, energy use costs $XX.XX compared to normal, off-peak hours when it costs $YY.YY. You can take advantage of the variable pricing and use less energy during peak hours, reducing your overall utility bill. By using less energy during peak hours, others have saved on average $5 to $30 per month.
With the north star in mind, I started hacking away. Does utility context convince customers to save energy? Not really. Do exact dollar amounts help? Probably not — you can just say it costs more or less. Does explaining why peak hours are peak hours help? Not likely; it’s also just context. Does social proof help? Yes, definitely, and it should be positioned in a way that leaves a lasting impression.
By the end, I landed with the version we shipped:
You’re currently on a rate plan that charges different prices throughout the day. By using less electricity when it’s most expensive, others have saved on average $5 to $30 a month.
This version gives the most important contextual details around what this email is, and the social proof to encourage further reading and action.
The rest of the report pairs clear, minimal messaging with data visualization to put this information in the context of their own energy use. I use simple, subtle prompts to nudge them further in the direction we want to move.
What parts of your routine could you do at off-peak times?
Final deliverables
Results & impact
The core goal of this program was to shift customer energy use outside of peak hours, reducing demand on the grid during the highest-usage hours. So the question is, did this program achieve that? Here’s what was reported:
The program delivered a 1–2% peak load shift across five utility deployments, with digital engagement reaching 60% among recipients at one utility.
As far as the content goes, one of the utility deployments with roughly 800,000 households moving through a time-of-use rate transition saw 3× the amount of email opens and 9× the amount of clickthroughs compared to the utility-industry benchmarks.
The email series also improved comprehension of their utility’s TOU program, with recipients being 17% more likely to recall the program and 16% more likely to correctly identify their peak-pricing hours compared to a control group. For an energy-shifting program, that level of comprehension is what you’d actually need to make behavior change possible.
Over time, the program saw an 8% lift in customer satisfaction versus control, and a 5% increase overall. We also saw 67% of customers noting a preference for the weekly cadence, which is a notable departure from the bill-cycle rhythm typical of utility communications. While it’s generally pretty rare to hear that customers actually want more emails from their utility, it’s a signal that more frequent utility contact can be a good thing when it’s done well and provides real customer value.