Incident management, and what most teams actually need
The category grew. The problem mostly did not.
Ask what incident management software does today and the answer is a list: on-call scheduling, alerting, status pages, incident roles, comms templates, postmortem workflows, a service catalog, dependency mapping, and lately an AI that summarises the incident while it is still happening.
Ask a four-person team what went wrong last time and the answer is usually one of two things. Nobody was woken up. Or somebody was woken up and it was the wrong person.
Both of those are paging problems
Neither is solved by a status page, and neither is solved by a better postmortem template. They are solved by a rotation somebody maintains, an escalation chain that gives up on one person and moves to the next, and a notification loud enough to survive a phone on silent.
That is a small amount of software. It is not a small amount of care, which is a different thing, and it is why the small part is worth doing properly rather than treating as the free tier of something larger.
What the rest of the category is for
Genuinely useful, when you have the problem:
- Status pages matter when you have enough customers that telling them individually does not scale.
- A service catalog matters when nobody can hold the list of services in their head.
- Enforced postmortems matter when incidents are frequent enough that the lessons get lost, and when enough teams are involved that the process has to be written down rather than remembered.
- Incident roles and comms templates matter when the response involves people who are not engineers.
Every one of those is a "when you are bigger" problem, and every one is sold to teams who are not.
The cost of buying early
Not just the invoice. A platform has a shape, and adopting it means adopting the shape: fields to fill, statuses to move things through, a process that assumes somebody owns the process. For a team of four that overhead is not neutral, it is a tax paid on the worst night of the quarter.
The specific version we hear most: per-seat pricing makes the rotation smaller. Teams quietly decide somebody does not need an account, and then the same three people carry every night. The tool did not cause the burnout, but it priced the fix.
What Acked does
- On-call schedules with layers and overrides.
- Escalation that does not give up until somebody answers, and tells you when nobody did.
- Paging that gets through: critical alerts on iOS and Android that ring past silent and Focus, plus email.
- Alerts from anything that can POST JSON, with first-class formats for Datadog, Prometheus Alertmanager and Grafana.
- A full history: every page, every delivery attempt, every ack, with who and when.
And a flat price with unlimited users, so adding the fifth person to the rotation costs nothing.
When this is the wrong answer
If you need status pages your customers read, a postmortem process the tool enforces, a service catalog, or incident roles and comms templates, buy a platform that has them. You will be happier and it will be cheaper than assembling the same thing out of parts.
Acked is for the case where you looked at that platform, realised you would use one tenth of it, and could not find anything smaller that was still serious about waking people up.