A Story of Bad Design During COVID-19

Two friends of mine, who also happen to be sisters, recently told me a design horror story they experienced first hand that I’d like to share. The story is about an app, designed by a country’s government, released recently to help people figure out whether or not they needed to be tested for COVID-19, but a UI design oversight led to confusion, frustration, and fear. As a designer hearing their story, I was horrified! I mean, one app, released during a pandemic, to serve an entire nation (we’re talking millions of people), resulted in outcomes that were the complete opposite of the app’s intention because of one design hiccup.

For the sake of privacy, the people, government, and mobile app mentioned in this article have been kept anonymous. The names of all involved have been replaced with pseudonyms.

Gina’s mom, Maya, downloaded an app, COVID Control, that was recently built and made public by the Ministry of Health. COVID Control has a self-diagnosis questionnaire that you can fill out to see if you are at risk of having COVD-19.

Maya was curious and took the test. She selected ‘no’ to all of the questions because she wasn’t feeling any symptoms and hadn’t traveled anywhere recently. Maya was a bit skeptical about the toggle and doubted for a second on whether she was selecting ‘yes’ or ‘no’. She then went with her gut feeling and submitted the test.

Maya was shocked to see that she was considered high risk of having COVID-19. She knew immediately what went wrong and rushed to tell her family. They all found it absurd and laughed it off.

The next day, having already forgotten about the app, Maya received a call. It was from the Ministry of Health, letting her know that she’s been identified as a high-risk patient and needs to get tested immediately. She was told that she will be taken to a room at the hospital where she will be quarantined as she waits for the test.

Maya didn’t expect this. She explained to the government official that there has been a mistake and that she meant to select no to all the questions. The official responded with, “you’re not the first call of the day to say you answered them incorrectly.” Despite this Maya was still told that she had to come and get tested to clear her name. The official and her team were concerned that people would start using inaccurate data as an excuse to avoid getting tested.

My friends were furious. They weren’t okay with sending their mom to a hospital at a time of a full lockdown where everyone feels the most safe in their homes. They feared she would be at risk of catching the virus if she were to go.

After some negotiation, they agreed with the official that their father would go to the hospital, take his wife’s national ID card with him and explain the situation. This would supposedly clear her name. As the father explained it to the doctor, the doctor also mentioned the fact that “you’re not the first person that comes to me and tells me you answered the question wrong.”

The app was made public by a press conference. This press conference happens every week, where the deputy of the Ministry of Health announces new COVID cases and any other updates regarding the precautions the country is taking, like the lockdown and curfew. You can imagine that the entire country is waiting to hear the news.

It was evident that this small UI component, the toggle, brought confusion to many, many people.

Human error when using a product or service is very common, hence as a product builder, you need to foresee that and design for several edge cases. A well designed user experience should prevent disasters like these from happening. Below are two common design principles I’ve surfaced that can help avoid similar disasters when designing products.

Design Principle #1: Identify if the action has a high-stake outcome and design accordingly.

Not all actions do, so they will have different treatments. In this case, the action had high stakes for a few reasons:

• If answered incorrectly, the user would receive inaccurate results on a serious topic, potentially causing confusion and fear.
• If answered incorrectly, the Ministry of Health would take actions based off of inaccurate data, resulting in serious consequences.

When designing a form, the answer selection needs to be clear and obvious. Now, if this form is for a mass audience, many factors need to be considered to make sure it’s accessible to users with different capabilities and needs, including taking the style, colour, and size into consideration. The yes or no selection should use a design that is universally understood.

If the results from the self-diagnosis test were just shown to the user, selecting the submit button and executing that action immediately would be fine. However, it turned out that these results were collected and recorded by the Ministry of Health. In that case, users should get prompted to confirm their answers before submission.

Design Principle #2: Be transparent and set clear expectations on how users’ information will be handled.

It should be made clear to users, before filling out a form, that their results will get recorded and they should be given next steps on what to expect. I had asked my friend to confirm whether this was made clear to them. She answered:

“It was not made clear in the press conference, nor on the app that the results submitted from the questionnaire would be stored in a centralized system, we just thought it would be a tracker.”

The app had users signing in with their national ID. For my friend, this made it official, and she was able to draw the conclusion that her results could be stored. However, for the average user, that isn’t as blatantly obvious as it should be.

When designing a form, the results that are given after the user submits their test should be accompanied with any next steps that the user should expect, as well as guidance on what they can do.

Conclusion

I called my friend the next day to let her know that I’d like to share her story and had a few questions to ask. Gina was happy to fill out the questionnaire herself so she can screenshot and share a few screens with me. Although, she was a bit skeptical to submit the self-diagnosis because she didn’t want to experience that chaos for the second time. She was extra careful with selecting ‘no’ to the answers since she too wasn’t experiencing any symptoms and hadn’t recently traveled.

Phew, she submitted it properly and her results came out as low-risk. We were curious to see what her mom received when she submitted her test, answering ‘yes’ to all the questions. Gina called her mom to casually ask her, and for that reason explained to her that she had filled out the questionnaire for me. Her parents reacted frantically to Gina downloading the app and feared going through that again. It took Gina some time to calm her parents.

After a while, Gina called me back. “They’re so traumatized.” She said.

It occurred to me that the app caused trauma and additional fear to many people who had used the self-diagnosis test. The intent behind a product like this one is to reach a mass audience and therefore make a significant impact. But if people are leaving unhappy, it is going to be difficult to lure them in again. And now, if a chunk of people are not using the service, how effective will it really be?

What started as an attempt to reduce the pressure on the healthcare system, only resulted in additional demand of resources at hospitals, increasing the risk of COVID-19 by exposing potentially ‘healthy’ people, and a loss of trust from the people. For a country that wants to advance and make use of digital services, gaining the people’s trust is critical, and therefore the design and experience of a service holds a lot of weight in making that happen.