Monitoring
Status Pages
Pricing
In the previous posts, we’ve looked at how alert noise emerges from design decisions, why notification lists fail to create accountability, and why alerts only work when they’re designed around a clear outcome.
Taken together, these ideas point to a broader conclusion. That alerting is not just a technical system, it’s a socio-technical one.
Alerting systems encode assumptions about how people behave, how responsibility is distributed, and how decisions are made under pressure. When those assumptions don’t match reality, the system fails; even if the underlying technology works exactly as designed.
Every alerting setup reflects a set of beliefs, whether they’re explicit or not.
Who gets alerted implies who is trusted to act.
How many people are notified implies how confident the organisation is in its ownership model.
What information is included implies what the system believes matters in the moment.
When alerts are broadcast widely, the system is effectively saying: we’re not sure who should handle this. When alerts contain excessive context, it’s often because the system doesn’t trust the recipient to know what to do without it.
These are not technical shortcomings. They are organisational signals.
Many alerting systems evolve as a response to uncertainty.
A missed incident leads to wider notifications.
A slow response leads to more context.
An unclear handoff leads to additional escalation paths.
Each change is rational in isolation. Over time, however, the system begins compensating for gaps elsewhere, whether unclear ownership, fragile trust, or ambiguous expectations, rather than addressing them directly.
The result is a system that interrupts many people, moves slowly, and produces stress without improving outcomes.
It’s tempting to treat alerting problems as tooling problems. Adjust thresholds. Add integrations. Create more rules.
Those changes can help at the margins, but they don’t resolve the underlying issue if the design assumptions remain unexamined.
No amount of configuration can make up for uncertainty about who owns a problem.
No escalation policy can substitute for clarity about what action is expected. And no alert volume reduction will help if alerts are being used to achieve outcomes they weren’t designed for.
The hard work here isn’t technical. It’s conceptual.
Good alerting systems are designed with a clear understanding of how people actually behave during incidents.
They assume limited attention, incomplete information, and hesitation under ambiguity. They reduce the need for interpretation and make the next step obvious.
When an alert fires, the recipient shouldn’t need to ask:
If those questions arise, the system has already failed them.
Over time, mature organisations tend to converge on similar alerting characteristics.
Alerts are fewer, but more trusted.
Ownership is explicit, not implied.
Visibility is provided through the right channels, not forced through alerts.
And responsibility is designed into the system, rather than inferred socially.
This isn’t about perfection. It’s about alignment between technology, process, and human behaviour.
Alerting systems work best when they reinforce clarity, not compensate for its absence.
Alert noise isn’t an accident. Notification lists aren’t teams. And alerts without outcomes are just interruptions.
Seen together, these aren’t isolated problems. They’re symptoms of systems that haven’t been designed with humans in mind.
Treating alerting as a socio-technical system doesn’t make the problem simpler; but it does make it solvable. Because once design assumptions are visible, they can be challenged, refined, and improved.
And when that happens, alerting stops being a source of stress, and starts being what it was meant to be all along: a system that helps people act, clearly and confidently, when it matters most.
Share this
7 min read A website may be standing and still be in trouble. It may answer a request, return a cheerful 200 OK, and yet load slowly enough that visitors begin to lose patience. Its certificate may be nearing expiry. Its domain records may have changed. A server may be filling its disk in the background, patient and
6 min read StatusCake tells you that something might be broken. Hermes can check whether it really looks broken, decide who should hear about it, send the email, and keep the record for tomorrow morning’s summary.
3 min read The allure of OpenClaw is undeniable. You deploy a highly autonomous, self-hosted AI agent, give it access to your repositories and inboxes, and watch it reason through complex workflows while you sleep. It is the dream of the ultimate 10x developer tool realized. But as any veteran DevOps engineer will tell you: running an LLM-backed
7 min read There are cloud outages, and then there are us-east-1 outages. That distinction matters because failures in AWS’s Northern Virginia region rarely feel like ordinary regional incidents. They tend instead to expose something larger and more uncomfortable: too much of the modern internet still behaves as though one place is an acceptable concentration point for infrastructure,
7 min read Artificial intelligence is making software easier to produce. That much is already obvious. Code that once took hours to scaffold can now be drafted in minutes. Boilerplate, integration logic, tests, refactors and small internal tools can be generated with startling speed. In some cases, even substantial pieces of implementation can be assembled quickly enough to
10 min read Whilst AI has compressed the visible stages of software delivery; requirements, validation, review and release discipline have not disappeared. They have been pushed into automation, runtime and governance. The real risk is not that the lifecycle is dead, but that organisations start acting as if accountability died with it. There is a now-familiar story about
Find out everything you need to know in our new uptime monitoring whitepaper 2021
