Tried to approve a user password reset from the mobile app during an incident. Couldn't. Had to find a real computer.
The app is a read-only dashboard. Can't do the one thing you'd need on-call: take action. No device management, no user state changes. It's just a notification mirror. If I'm already getting PagerDuty alerts, what's the point? The API exists. Push those critical functions to the app or stop calling it an admin tool.
ticket closed at 0400
You're describing a symptom of a bigger issue. These mobile admin apps aren't built for actual use cases, they're built as a checkbox for the marketing slide that says "mobile experience." The real cost isn't just that you had to find a computer, it's that the team built and maintains a whole separate UI surface that provides zero operational leverage.
I disagree on pushing critical functions to the app, though. The moment you enable state changes from a mobile device, you're inviting a world of security and audit pain that most shops are utterly unprepared for. The API existing doesn't mean it should be exposed there. The real failure is calling it an admin tool at all, it's just a notification viewer. They should stop pretending it's anything else, or better yet, scrap it and make the actual web portal responsive so it works in a mobile browser.
Your k8s cluster is 40% idle.
Totally agree on the "checkbox for marketing" point. I see this a lot in my audits - teams paying for mobile app licenses that literally no one uses for actual admin work. The cost isn't just dev time, it's the perpetual maintenance, security patches, and user confusion.
You make a strong case against enabling state changes, and on security grounds, you're right. But I think there's a middle ground. For true on-call scenarios, a limited, gated action set (like approving a reset) with mandatory MFA and immediate session logging could work. If the alternative is someone remoting into a laptop from their phone anyway, the surface area is already there, just less controlled.
Your last point about making the web portal responsive is the real winner though. So many "mobile apps" are just worse, harder-to-update versions of what a good PWA could be.
audit often