Lessons from the COVIDSafe App: Why Public Health Tracking Apps Fall Short

When Australia launched the COVIDSafe app in 2020, it aimed to automate contact tracing using Bluetooth to detect close contacts. The idea sounded promising—real-time alerts, faster outbreak detection, fewer manual efforts. But in practice, the app struggled from the start. Users quickly lost trust because it wasn’t clear how their location data was stored, accessed, or used. The government didn’t explain the boundaries of data access, and without transparency, people stopped installing it. By April 2020, the app had just over 7.7 million downloads—less than 3% of the population. That’s not just a low number; it’s a sign that even a well-designed app fails if people don’t believe in it or find it hard to use.

The app had serious technical flaws too. Registration was clunky, location tracking often failed, and it didn’t work reliably across devices. Worse, it only flagged contacts when people were within a certain range—so it missed many real interactions. That meant it couldn’t catch outbreaks early or provide a full picture of transmission. It also didn’t connect with existing public health systems. Contact tracers couldn’t pull data from the app to support their work, making it a ghost in the machine. And when the app was eventually shut down, users were left confused—especially because the original plan included a sunset clause, meaning it would end after a set date. That lack of clarity made people wonder what happened to their data and whether it could be used later.

Key Lessons from the COVIDSafe Experience

• Data privacy and transparency matter more than ever: People won’t trust a public health app if they don’t understand how their data is handled. Clear rules, plain language, and real-time accountability are essential.
• Technical flaws hurt adoption—no matter how smart the design: An app that crashes, doesn’t work across devices, or requires too much effort to use will never reach the public. Usability must come before functionality.
• Passive tracking is not enough: Relying on Bluetooth to detect proximity misses most real-world contacts. Public health tools need to combine digital signals with active, human-led tracing to be effective.
• New apps must fit into existing systems: If data doesn’t feed into current workflows, it’s just another isolated tool. Integration with contact tracing teams and databases is non-negotiable.
• Sunset clauses don’t fix data uncertainty: Planning for how data is stored and used after an app ends is just as important as building the app. Without it, users feel left in the dark.

Public health apps can help—but only if they’re built with people in mind, not just technology.