Orchestrated Energy (OE) is a program launched in the summer of 2016. It works to optimize residential HVAC runtime during summer months by integrating with smart thermostats. Utilities looked to OE to help solve critical issues they face with grid stability and energy efficiency.

During my time on OE, I was part of a cross-functional team that delivered the next generation of energy efficiency through smart home technology.

Challenges & Objectives

OE functions on two modes, the first being Energy Efficiency mode, which creates personalized daily energy efficiency schedules for individual homes. The second is Demand Response mode, which responds to peak energy demand days by optimizing thermostats within a given territory to shift the energy demand to different, less expensive times of the day. During my time on the project, OE ran on Nest and Ecobee smart thermostats.

OE is also an opt-in software, meaning on a day-to-day basis, users could simply adjust their thermostat, called an override, thereby disrupting Energy Efficiency mode or opting-out of a Demand Response event entirely. In order to maximize energy efficiency for both the utility and end-users, ideally OE would operate the majority of the time uninterrupted. In order for utilities to meet their peak energy demand targets, it was imperative that end-users not opt-out of an event.

My Role

I led the design of the OE customer experience from January 2017 until April of 2019 as part of a cross-functional team of engineers, data scientists, building model experts, product managers, UI designers and marketing experts.

My day-to-day responsibilities included conducting user research, wireframing, prototyping, designing flows, UI design, usability testing and workshop facilitation.

Kick-off & initials concepts

When I joined the team, OE had already operated a pilot program for a full summer. I partnered up with our customer success and product teams to quickly understand early key learnings:

1. Users want to know what OE is actually doing to their thermostat

2. Users want to know why OE is doing what it is doing

3. User want to know if OE is working

Onboarding & Education through MyHome

Majority of users’ pain points circled around their lack of knowledge about what OE was doing, therefore we focused most of our efforts on providing a rich onboarding experience supported by intermittent engagement through the use of Tendril’s app, MyHome. MyHome was developed by another team to be an “app of apps” whose function was to integrate all of the resources an end-user would need from their utility company into one simple place.

Working within the existing constraints of MyHome, we leveraged its Newsfeed feature to guide users through installing OE on their smart thermostat. I collaborated with a product manager and copywriter to help design the flow of dialogue and interaction points for the user.

In addition to leveraging the Newsfeed feature, we also built a mini-thermostat dashboard to give users a visual indicator as to what was happening on their thermostat. This also helped us get around user interface restrictions we faced with our smart thermostat partners and their devices. The Thermostat feature initially displayed the actual temperature, the setpoint and the OE modes; Learning, Active, Demand Response and error states. The green and red rings had unique animations that corresponded with the various modes. Many iterations of the dashboard were sketched, designed, prototyped and tested.

I worked with product to design and scope additional, more detailed views of what was happening in individual users’ homes. However, due to limitations with data availability and resource constraints, the designs were never implemented.

Sunsetting MyHome

At the same time the team was uncovering issues, MyHome was also experiencing slow growth and, due to a unique sales strategy, utilities were opting to purchase OE without MyHome. Without a clear sales pipeline for MyHome, resourcing became a constraint. Scaling it to meet the growing needs of OE users became unsustainable and they were clearly feeling the impact. The app rarely updated, the engagement strategy lagged far behind, and enhancements that would provide more insight and control were tabled. With no roadmap or resources, the team had to make the tough decision to sunset MyHome, even though it meant losing our only method of engaging with users. We were back at square one.

Bootstrapping the Customer Experience

The resourcing pains we felt around MyHome were universal. Building any type of digital platform to replace MyHome was off of the table. To solve the issues we were uncovering and meet our contractual obligations, we would have to get creative. 

First, in order to address users opting-out of Demand Response events and overriding Energy Efficiency, I knew we would need to address it at the root cause—uncomfortable temperatures. 

I pulled together a small team consisting of myself, a data scientist and a market researcher. Together we compiled all the research we had gathered on our users and created four Comfort Personas.

Next, I collaborated with a product manager and led a team of engineers, data scientists, and building model experts through a design sprint-like workshop. Using our comfort personas, we created journey maps of our users’ experiences on various days. We identified key moments of the experience that were driving users to disengage. We brainstormed and storyboarded potential solutions to test. The top ideas all involved engineering work to improve functionality of smart features and science work to improve algorithms. 

Due to the proprietary nature of both, I cannot detail them further. However, the results spoke for themselves. In subsequent tests of the improvements, the team was able to reduce the number of too-hot overrides per day by roughly 50%.

While the engineers and scientists worked to improve comfort, I began to collaborate with our marketing expert and product manager to strategize a content and engagement plan. We invited our utility stakeholders on-sight to participate in a 2-day workshop. I walked them through similar journey mapping exercises and brainstorm sessions.

Using the insights and ideas from our workshop, we conceptualized a series of emails that would help to replace the Newsfeed experience. These emails would be content rich, filled with educational material, home performance metrics, tips and surveys. And, they would be timely and relevant to what was happening in a user’s home. I created very low-fidelity wireframes and handed them off to be designed by a visual designer.

We knew relying on email had it’s disadvantages. There would be a lot of competition for users’ attention in the inbox. Emails also couldn’t deliver the real-time information that the Thermostat dashboard could. Still, we hoped that by timing sends after key events like a Demand Response event, or sending a monthly report at a regular cadence, users would feel informed and rewarded. Plans to send important messages in SMS were also put in place.

The gamble paid off, our email open rate was extremely high and users who recalled receiving the email report greater overall satisfaction with the program.

Outcome

OE is a complicated and nuanced product with as many external dependencies as internal ones. Though the team faced significant obstacles, we were still able to deliver a positive experience through iteration grounded in research and human-centered methods.