Posts from "January 2013"

Our Thinking

Engagement first: the key to developing utility mobile apps that create business value

  • By Jordan Jakubovitz
  • January 28, 2013

Almost a quarter of our blog readers access our site from their mobile devices. This isn’t too surprising, since more than half of mobile subscribers in the US have a smartphone. Similar mobile adoption trends are taking shape around the world. And it’s not just that people have smartphones; it’s that they use them. A lot.

That’s why today we are announcing Opower Mobile. This new app will help our utility partners engage their customers in ways that are meaningful for customers and beneficial for utilities. By bringing the increasingly popular Opower 4 platform to smartphones, utilities can empower customers to take control of their energy use anytime, anywhere on their mobile device. And because the app is easily customizable, utilities have the ability to deploy it in a way is tailored for their needs.

The potential of mobile

It’s easy to see the business potential in mobile. Mobile apps are changing the way we communicate, shop, and learn about the world. They’re also changing the way we interact with traditional service providers. Over the next five years, almost half of U.S. bank account holders will be banking through mobile devices. Searching for “banks” in the Google Play store returns results for almost every major provider of financial services. Other service-oriented industries are adopting similar strategies, and for good reason: engaging customers through the mobile channel represents an opportunity to differentiate from competitors, increase customer satisfaction, and transition customers to self-service (mobile billing, for example).

Source: Accenture (2010)

The same motivations are true in the utility industry. Utility companies can realize significant benefits by empowering customers via self-service. And in fact customers are increasingly demanding these new and innovative tools from their utilities. In response, many utilities are beginning to introduce mobile apps to the market.

The utility mobile landscape

Utility applications currently follow two distinct development philosophies. Some are “transaction-first” applications, built to help a customer complete a transaction from a mobile device (reporting an outage, for example). Others are “engagement-first” applications, built to encourage and maintain user adoption. While those applications still include transaction capabilities, they prioritize user experience and interactive discovery.  We’ve observed that about 90% of utility applications now in-market take the transaction-first approach.

Transaction-first applications tend to prioritize user completion of narrow tasks such as bill pay. When rolling out these  kinds of offerings, utilities most commonly use off-the-shelf development frameworks that are quick to implement but functionally limited.  In this way, transaction-first applications run the risk of creating an unpleasant association between the mobile application and the primarily neutral-or-negative user experience that it delivers – such as paying a high bill or learning of a service interruption. As a result, these applications typically have 2-star ratings in the app stores and low rates of adoption. (Compare that, for example, to USAA’s 4.5-star app.)

Engagement-first applications present a different approach to meeting business objectives. Rather than focus on narrow and sometimes unpleasant transactions (e.g. bill pay), these applications prioritize features that drive positive user experiences. This creates a strong foundation on which to layer transactional elements. Utilities with successful engagement-first applications have taken a cross-channel approach to marketing their apps and prioritizing discoverability. They correctly balance features that meet core utility business objectives with features that delight customers. Utilities that have taken this approach have achieved 4.5-star-rated applications that rival the best of the more successful financial services applications.

The challenges of engagement-first approaches

It’s not easy to build a successful engagement-first application. The first reason is that mobile users have diverse products and operating systems (see the colored pie chart below), and there are indications that  the  mobile ecosystem is going to get more complex. In addition, we must also acknowledge that a mobile user isn’t only a mobile user. It’s important that customers have a consistent experience across all channels (web, mobile, tablet, and whatever might come next). And we can’t forget the crucial emphasis that engagement-first applications have to place on design principles, since usability is vital to engagement.

Smartphone Operating Systems

Creating a seamless front-end experience is just one piece of ensuring a successful engagement-first app. Utilities must also consider the content in the application: beyond transaction features, what do customers actually want? Is it consistent with their experience across channels? What will motivate them to return to the application frequently? And how can utilities best provide those features? Our research has revealed that, among other things, customers seek usage information, energy analysis, and device controls. Successful engagement-first apps must be supported by a back-end infrastructure that can deliver this content, grow with new customer demands, and enable a customer relationship that’s coordinated and personalized across channels.

Lastly, mobile apps add the most value when they provide better visibility into customer actions and preferences. It’s essential to set up the proper instrumentation and reporting so that utilities can continually measure engagement and make the right adjustments over time. Responding to customer feedback (both vocal and behavioral) is a key ingredient to the success of any mobile app.

Keeping up will be hard, but it’s essential that engagement-first applications do so. Through the right mix of turnkey availability and customization opportunities, utilities have a chance to simultaneously meet their own needs and delight their customers.

Opower’s solution

At Opower, we’re focusing on building a great engagement-first mobile application (see more details here). We’re confident that this app will, as an extension of the broader Opower 4 platform, help utilities meet current customer expectations and provide a future-proof way to take advantage of new advances in mobile technology. We look forward to hearing your feedback on our vision.

Jordan Jakubovitz is a member of Opower’s Marketing and Strategy team.

Read More

Will the Super Bowl save the planet? How America’s most watched TV event reduces home energy usage

  • By Barry Fischer
  • January 27, 2013

The 2012 Super Bowl was the most watched television broadcast in US history. An estimated 111.3 million viewers tuned in – more than a third of the country’s population.

That’s a lot of TVs aglow at one time.  The amount of electricity collectively consumed by TVs during the big game is reckoned by General Electric to exceed 11 million kilowatt-hours – equivalent to the power generated by ten medium-sized coal-fired power plants over the course of the event.

But here’s the real kicker: all those TVs illuminated during the Super Bowl are actually a force for dramatically lower overall energy consumption.

Turning back the clock to February 5th 2012, we analyzed electricity consumption data across 145,000 households on last year’s Super Bowl Sunday. And our analysis revealed a remarkably consistent pattern: when the game kicks off, electricity usage plummets.

Decrease in US home electricity use is 3x the amount of energy consumed by the TVs watching it

Let’s examine the game-day details and explore how a major televised sporting event can affect home energy use…

When Super Bowl XLVI kicked off, home electricity usage dropped off

First, let’s take a look at game-day energy usage in the western part of the country, for which we analyzed data from 91,000 anonymous households.

The chart below shows how average residential electricity consumption on 2012’s Super Bowl Sunday (broken up into 15-minute intervals) differed from a typical midwinter Sunday. From morning through midafternoon, Super Bowl Sunday was a fairly normal energy day in the West.  Between 10am and 3:30pm, electricity usage was slightly above typical – possibly as party hosts were doing preparatory tasks like washing dishes, and soon-to-be party guests were firing up their ovens to bake jalapeño poppers. 

But then something special happened around 3:30pm PST: the Giants and Patriots ran onto the field at Indianapolis’ Lucas Oil Stadium. Immediately, electricity consumption began to tumble downwards. As the game’s suspense escalated, usage levels descended –dropping to around 7% below what is normal for a Sunday afternoon/evening.

During the halftime show, the game’s viewership surged to 114 million people (even higher than the 111.3 million that the game averaged), who were treated to performances by Madonna, the dance duo LMFAO, and the English singer M.I.A.

But something else also seemed to be M.I.A. during halftime: home electricity usage. It faded to a full 7.7% below typical levels.

And so while in the hours before the game, home electricity usage had been just slightly above (+2.4%) what is typical, things changed radically after kickoff. Over the course of the game, consumption decisively decreased to 5% below normal.

And usage remained well below average (-3.7%), even after the game was over.

Why do homes use significantly less electricity during the Super Bowl?

We think there are a couple key reasons for the sharp reduction in usage:

1) Many households are not using appliances other than their TV during the game. On a typical Sunday in the early evening, many people would have been at home using multiple different kinds of electricity-consuming appliances (e.g. laundry machine, kitchen appliances, vacuum cleaner, TV) around the house—as suggested below by the relatively high evening usage that normally takes place on a Sunday.

But the Super Bowl changes this pattern: it concentrates activity around the TV. With so many people glued to the couch during the game, fewer households are using electricity for cooking, cleaning, or anything else other than watching the tube. Abstaining from other forms of energy usage naturally causes game-time electricity consumption to decline below typical levels.

And what explains the persistence of below-average consumption levels even after the conclusion of the game? One factor (another is described below) may simply be sustained fixation on the TV, which continues to divert people away from their typical Sunday night uses of electricity.  For example, after the game ended in 2012, tens of millions Americans didn’t leave the couch: 37.6 million viewers stayed huddled around the TV to watch the season premiere of the musical reality-TV show “The Voice.”

2) Many people watch the game at houses of friends and family. For many fans, watching football is a community event — meaning that people leave their homes to watch the game with other fans. In 2011, an extensive poll by Nielsen found that 45% of Super Bowl viewers planned to watch the game with friends or relatives.

A mass movement toward collective TV-watching at friends’ houses (or at a bar) on a Sunday night will result in significantly lower-than-average electricity use. Just as carpooling reduces transportation energy use, gathering together to watch televised sports—let’s call it “TV-Pooling”—decreases home electricity use. Twenty people watching one large TV at a friend’s house requires much less energy than 20 people watching 20 TVs in their individual homes.

And because for West coast audiences the Super Bowl concludes fairly early (around 7pm), the communality of the game-watching is likely to segue into additional away-from-home social revelry that persists longer into the night…as is suggested by the sustained lower-than-average usage through the very end of Super Bowl Sunday.

The numbers and analysis above correspond specifically to the Western region, but what about elsewhere? To test whether the Super Bowl’s energy-saving effect truly represents a broader nationwide phenomenon, let’s check to see if the pattern we’ve identified also materializes on the other side of the country.

Chips and (Electricity Usage) Dips: across the country, home energy consumption declines during the Super Bowl

Now we know what to look for, as we shift our attention to 54,000 anonymous households located in the eastern part of the country…

When last year’s Super Bowl kicked off (at 6:30pm EST), did the region’s average home electricity usage substantially decrease like it did in the West?

Indeed it did: game-time consumption in the East dropped to as much as 5% below typical levels.

Up until game-time, usage in the Eastern region was slightly below normal — likely due to the relatively warm weather that day, which may have given people an incentive to spend more time outside.  But then the Giants and Patriots took the field in Indianapolis, and poof: electricity consumption decreased significantly. Each 15-minute interval of usage during the heart of the game was more anomalously low than any other interval that day.

Consolidating the 15-minute usage intervals into wider time periods, we see that home electricity usage during the game was, in total, nearly 4% lower than we would expect on a typical midwinter Sunday evening – and represented the day’s most significant deviation from normal.

The pronounced decline in game-time electricity usage in the East is, similar to what we saw in the West, most likely a consequence of two related behavioral phenomena: exclusive focus on the TV and communal game-watching.

Importantly, these two subconscious energy-saving actions take place during a time that is normally characterized by peak residential consumption — i.e. when people are typically at home doing their Sunday evening routine (cooking, washing, etc.).  It’s for this reason that the Super Bowl produces a prominent energy-conservation effect: relative to a run-of-the-mill midwinter Sunday evening, the Super Bowl catalyzes an extensive cutback in household energy use.

And why did eastern residential electricity consumption spike to above-average levels after the game, given that the entire rest of the day had been below average? Quite simply, the Super Bowl ends on the late side in the Eastern time zone. That means that immediately after the game, many people (at least those planning to report to work the next day) were probably returning home en masse from parties.  One can imagine that as they walked in their doors, they collectively flipped on the lights and other appliances. This, in turn, translated into an unusually above-average period of energy usage relative to a typical Sunday’s late night, when most people would have been winding down and going to bed.

Better together: TV-Pooling is good for the planet, the pocketbook, and civic life

We were struck by how the Super Bowl – which has for the last 3 years been the most watched television event in US history – appears to reduce home energy usage.

Mass TV watching traditionally conjures up images of enormous, electricity-gobbling 42-inch plasma screens.  But as it turns out, all these TVs are a force, at least over the course of a few hours, for decreased electricity consumption.

And we were especially intrigued by the energy-efficiency implications of a key factor behind the decrease: TV-Pooling.

Communal TV viewing may at first seem like a trivial concept, but its effect on a country’s residential energy consumption could be significant. Super Bowl XLVI demonstrated that when around one-third of Americans collectively watch a single 3.5-hour sporting event, the corresponding reduction in the nation’s daily energy bill can be upwards of $3.1 million. That’s a lot of guacamole. Replicate this phenomenon a couple times each month, and you’re potentially talking about some serious energy and cash savings.

In addition, TV-Pooling may do more than conserve energy: it may help strengthen our social bonds. In his influential book Bowling Alone, Harvard Professor Robert Putnam chronicles how Americans have become increasingly disconnected from their friends and neighbors in recent decades, replacing communal engagements with individualized entertainment. In Putnam’s analysis, television is among the culprits – and perhaps appropriately so. But, there is something different about how we watch major broadcasts like the Super Bowl.

Events like the Super Bowl are not just popular, they are social. And, as such, they may help, just a little, to unite neighbors, friends, and family – and help save the planet.

Special thanks to Ashley Sudney, Efrat Levush, Steven Blumenfeld, Nathan Srinivas, Emily Bailey, Elena Washington, Yoni Ben-Meshulam, David Moore, Katie DeWitt, Andrew Sharp, Nate Kaufman, Jordan Jakubovitz, Carly Baker, and Arkadi Gerney.

Follow @OpowerOutlier on Twitter


The aggregated data used for this analysis stems from two large regional samples (nwest = 91,355 households; neast = 53,574 households). Time-of-use consumption data in both cases are defined at an AMI granularity of 15 minute intervals.

Baseline electricity usage levels (used to evaluate the deviation in home electricity usage on Super Bowl Sunday, relative to a typical day) correspond to energy usage levels observed on other Sundays in January 2012 and February 2012 that exhibited similar regional weather to February 5, 2012.  Similarity in weather is based on comparable daily mean temperature and negligible-to-zero precipitation.  Each regional analysis is demarcated by a geographic area with maximum radius of 25 miles.  To reinforce the analytical comparability of daily weather conditions and the underlying electricity usage requirements associated with them, the analysis is explicitly restricted to homes that utilize gas heating systems. Meteorological data is sourced from Weather Underground.

Each 15-minute game-time usage interval exhibits a standard deviation of no more than 0.40 kWh. Accordingly, the computed interval differences from baseline usage are statistically significant at the 99% level, within an confidence interval of  +/- 0.005 kWh (equivalent to the amount of electricity consumed by a single 20-watt efficient light bulb during the interval considered).

Note that, in the western region case, the above-average usage levels observed during the pregame period are more than offset by the strongly below-average usage levels that register when the game starts. This suggests that Super Bowl Sunday does not simply reduce usage during a certain segment of day; it also, as a result of the exceptional game-time contraction in consumption, is an overall net reducer of daily residential energy use. This outcome holds true across both regional analyses.


TV power demand in terms of number of power plants: General Electric estimates that all TVs beaming the Super Bowl in the US collectively use 11,309,607 kWh over a 5 hour period, implying an instantaneous power demand of 2,262 MW.  Net capacity of an average coal-fired power plant in the US is 228 MW. Aggregate demand of 2,262 MW over the course of the event thus corresponds to approximately 10 average coal-fired power plants. This is a conservative assumption, as it does not reflect transmission and distribution losses, which are generally considered to be around 7%.

Game-time usage reduction > 3x the energy consumed by TVs watching it: Based on GE’s 5-hour TV energy consumption estimate of 11,309,607 kWh. Proportionality to our game-time-specific analysis suggests that all TVs watching the big game for 3.5 hours = (3.5/5) * 11,309,607 kWh = 7,916,725 kWh.  The larger of our two regional time-of-use evaluations suggest that average residential electricity usage falls to 5.0% below typical levels during the game. A given US household consumes an average of 11,496 kWh per year (i.e. 8,760 hours), and so for a 3.5 hour period is proportionally assumed to consume 11,496 kWh * (3.5/8,760) = 4.59 kWh.  Average per-household usage reduction during Super Bowl = 4.59 kWh * 5% game-time reduction = 0.23 kWh. Across 114.76 million households in the US, the specific game-time reduction nationwide sums to 26,355,811 kWh. This is approximately 3.3 times as much electricity as the 7,916,725 kWh consumed by TVs during the game. This is a conservative estimate, insofar as the Super Bowl takes place on a Sunday afternoon/evening, which is a relatively high-usage time of the week for the residential sector.

Game-time monetary savings almost equal to cost of a 30-second commercial: Specific game-time reduction sums to 26,355,811 kWh (see above). At a national average residential cost of electricity of $0.1187/kWh, the game-time energy cost savings is $3.13 million. 30-second spots during Super Bowl XLVI fetched $3.5 million.

Data Privacy: All data analyzed here are anonymous and treated in strict adherence to Opower’s Data Principles.

Author’s note: The analysis and commentary presented above solely reflect the views of the author(s) and do not reflect the views of Opower’s utility partners.

Read More
Our Thinking

Steven Blumenfeld, Come on Down!

  • By Hillery Brown
  • January 25, 2013

Steven Blumenfeld recently joined Opower as a Policy & Research Analyst on the Regulatory Team.  In this role, Steven will explore the energy policy landscape—both domestically and internationally—to help Opower thrive within existing markets and grow into new ones.

Prior to joining Opower, Steven worked as a consultant in the national security arena where he helped clients identify risks and opportunities through an analysis of government spending trends, shifting policies, and technological developments.  Through this role, he worked on issues ranging from DNA forensics to the counter-improvised explosive device market to supply chain security.

Steven also took part in internships at the U.S. Department of State’s Office of Environmental Policy in Washington DC and the Secretariat of the UN Convention on Biological Diversity in Montreal, where he focused on energy and water issues, respectively.  Steven completed his Bachelor of Arts at Yale University as a Political Science and International Studies double-major.

1. What made you want to join Opower?

My interest in energy policy began when I noticed solar panels in a hut in Swaziland and has taken me all the way South to explore the world’s only educational base in Antarctica powered entirely off renewable energy.   I was very excited to learn about the innovative and entrepreneurial steps Opower is taking to engage consumers about their energy use to help them save power, money, and emissions.  After learning a bit more about Opower and meeting other employees during the interview process, I was set on taking this leap to further pursue my passion for energy and environmental issues while working on something exciting, challenging, and meaningful.

2. What has been your favorite thing about Opower so far?

It is such a fantastic opportunity to know that the work I do each day is not only aimed at growing our business, but that in doing so we are also creating a fundamental shift in our energy system by empowering consumers with the information necessary to make informed decisions about their energy use.  In addition to the “double bottom line” nature of our company, it is awesome to work with such a talented and diverse group of people in a work environment that fosters collaboration, creativity, and fun!

3. What are some of your hobbies?

When not in the office, you can find me checking outthe many street fairs and new restaurants in DC, inhaling the chocolate milk at the Dupont farmers market each Sunday (while also stocking up on produce), and making the occasional day-trip for a new outdoor adventure!

Read More
Older Posts