“I’m not getting a connection. The app says it’s disconnected,” a panicked woman said to her husband, hastily showing him the app. They were newlyweds living in Manhattan, who had been building a second home in upstate New York as a quiet escape from the city. Every weekend they drove upstate to add more progress to their dream home. They had begun picking out furniture and had already connected smart home devices to monitor everything from the temperature of the pipes to motion happening outside their home. While in Manhattan during the week, they used this app to check up on their dream home.

“I don’t know, honey,” her husband answered, tapping on his own app. He force-quit the app and turned his own Wi-Fi on and off again to see if the screen changed. It still showed the same panic-inducing icon indicating no signal was getting through. A rush of stress and anxiety blanketed the couple. “Let’s drive up and check on it tomorrow,” he said.

With the upstate home miles away from any other house, there were no neighbors nearby that could check on it for them. They imagined the worst. A fire burned their home down to a crisp. A bear came and rampaged through the living room. A tree fell onto the house, causing it to crumble.

The next day, they made the three-hour drive upstate and arrived, to their relief, at a house that was just fine. They went in to investigate what caused the lost connection in the app and found that the Wi-Fi had gone out. In fact, the whole area had lost coverage for the past day.

Had they never installed these IoT devices, they would never have suffered the anxiety they experienced and rushed up earlier than they’d planned.

When Designs for Peace of Mind Cause Unintended Anxiety

The IoT experience we had designed was a mobile application connected to multiple IoT devices that was intended to give users peace of mind. We designed it to notify users any time something went wrong, such as if the indoor temperature got too low, the fire alarm went off, or the camera detected unusual activity. We used mobile notifications and email to get the user’s attention and the application UI to communicate the details.

We assumed that the ability to know these things would enable our users to relax and let the mobile app handle the worrying. But we were wrong.

Users ended up worrying more. They would constantly check the app and feel more anxiety about their house than they did before. When our application told them about potential issues, users would think about these issues all day.

One mistake we made in the design was that we sometimes sent too many notifications. When cameras detected any unusual activity—which might just be a shadow moving along the wall throughout the day—it would notify the user every few minutes. With so much noise in the communication, it caused more annoyance and anxiety than peace of mind.

Researcher as the Observant Wallflower

As the couple shared this story with me from their living room in Manhattan, I adjusted my camera to make sure it was recording correctly. I wanted to capture their crestfallen faces and the tone of their speech as they dove into this story. As a UX researcher in IoT, my role was to listen, record, and observe without injecting any of my own opinions or knowledge.

I’m often placed in this situation, where I’m an active listener, taking notes, watching a technological solution completely break down in front of me while I’m in a customer’s home, and not able to offer help or even give a nudge of advice. To keep the data clean, I try to become a wallflower so I can send the video back to the team and we can get working on a solution.

Still, my heart wrenches as I hear these anxiety-ridden stories and the cries for a better solution. The people I meet in these studies always have high hopes for their new smart home—their heads filled with expectations that were outlined in glossy marketing materials across internet ads and hardware store aisles—only to have them crushed by technological and user experience barriers.

In this scenario, where the Wi-Fi had gone out, I didn’t know of any magical solution that could have kept their anxiety at bay. I wished there were some secret section of the app I could show them that would have an answer—a way to prevent future anxiety from stressing them out—but it didn’t exist.

Instead I offered a sympathetic look, and said, “I’m so sorry to hear that this happened to you. I’ll have the team investigate and see what we can do for you in the future.”

The couple’s eyes lit up ever so slightly. Even though I couldn’t offer any direct help, oftentimes knowing that I can be a portal to the team that can take action is a big relief.

The Challenges of IoT

After packing up my camera and notebook and leaving the couple’s home, my coworker and I were finally able to debrief and speak our thoughts out loud.

“What did you think of the Wi-Fi scenario? Is there anything we could have done?” I asked.

He furled his eyebrows ever so slightly and shrugged. He began talking through the nuances of the scenario and we brainstormed out loud. If the Wi-Fi went out in the entire area, it meant the hub that connects the IoT devices to the Internet wouldn’t have been able to communicate to us. Our servers were getting no information, and as the app is currently built, it can only communicate that the devices are offline.

“Can we let them know that it’s just a Wi-Fi outage, and not that a cable burned down or something?” I asked.

“Not really. At this point in time we just wouldn’t hear anything back from the devices, and we wouldn’t know the reason why.”

One of the toughest things about building IoT products is accounting for the vast universe of factors that you have no control over. Everything from local Wi-Fi coverage, to the way a house’s electricity is wired can have a big impact on the overall user experience of the products. And since each home is different and there is technically no way to detect these nuances when a user is setting up a device, you’re left to almost guess blindly as to what the devices should say or do when users run into various issues.

Analyzing Anxiety

Back at the office, I spent the next few days working with the team to break down the interview footage into quotes and snippets of text to analyze. If there was no simple technical solution, we needed to at least delve into each scenario to understand what we could have done from a communication or design standpoint to ease some of the pain.

At this point, the walls were plastered with groupings of so many Post-it notes of quotes and themes that the Sharpie marker fumes were starting to get to us. Photos of how the IoT devices were set up in the research participants’ homes were taped to white foam boards surrounded by our hand-scribbled notes. On a large monitor, I was going back and forth between various video clips of the interviews, seeing if I could catch any moments that might lead to insight.

Principles for IoT

At the end of the analysis, we came up with a few principles for IoT:

Set realistic expectations. In the scenario with the couple and the home in upstate New York, they had anxiety when our offering couldn’t effectively monitor the safety of their home. But there were many moments of delight with much lower stakes, such as being able to show progress on construction. Setting the expectation that an IoT offering can be a fun way to capture interesting moments, but not promising completely safety or security, can go a long way toward ensuring that users don’t rely on the service for more than it can live up to.

Have levels of severity in communication. In the Wi-Fi outage scenario, since we weren’t equipped with enough information about why the connection with their device was lost, we could have made sure that the notification and/or screen doesn’t look too severe and communicates what we don’t know. Anything that may make a user feel like there’s an emergency, such as attention-grabbing red alerts, should be saved for scenarios where we have more certainty on how severe the situation is.

Keep communication human. Sometimes there’s an error, and, unfortunately, there is not much the technology can do to help. In those cases, keep the communication in the app or device as human as possible. It can go a long way in using sympathetic language that is responsive to the user’s potential emotional state at that point in time.

Account for external factors. Perhaps IoT devices can’t currently detect external factors such as the local temperature or Wi-Fi connectivity in the areas, but there may be APIs or other services that can be plugged in to to provide your devices more context.

Slowly Improving on IoT Experiences

Coming up with the principles was just one of the first steps toward reducing anxiety in the smart home experience. It takes time for this type of mindset to trickle throughout an organization and actually affect the real experiences of IoT products. For months I continued to conduct various in-home studies, hearing similar stories about both the joys and anxieties of living with a smart home. Over time, the stories of anxiety slowly started to wane as people lived with IoT products longer.

A few months later, an email arrived in my inbox from the couple with the home in upstate New York. They included a link to a video clip. I pressed play and watched as a bear started walking up to their home and rummaging through the bushes. My heart started pounding loudly as I held my breath and watched. Then, the bear slowly walked away.

“First bear sighting of the season. This was awesome!” the couple wrote in the email.

I breathed a sigh of relief and spent a few minutes calming down my own anxieties before getting back to my research analysis.