Three objects, three decisions
| Event | Metric | |
|---|---|---|
| What it is | One thing that happened | A number computed from events |
| Cadence | The instant it occurs | Recalculated daily, weekly or monthly |
| Decision it enables | Act on this one occurrence now | Judge a trend over time |
An alert is a rule layered on top of either one – "tell me now if this event happens" or "tell me now if this metric crosses a line".
15 metrics across six categories
Acquisition
Activation
Revenue
Retention
Reliability
Support
7 events worth seeing in real time
Alert on these, not on every metric move – alerting on everything just trains people to ignore alerts.
Two starter dashboards
Pre-PMF
Signups, activation rate, and every support request – you're still learning who your users are and what breaks for them.
Growing SaaS
MRR, churn rate, failed payments and job failure rate – the metrics that catch revenue and reliability problems before they compound.
Frequently asked questions
Should every metric have an alert?
No. Alert on the 7 events above and on hard thresholds for 1-2 metrics you truly can't miss. Everything else belongs on a dashboard you check, not a notification you get.
Can LogsNinja compute metrics from events automatically?
Yes. Send the underlying events with the right metadata, then build a metric or chart on top – count, sum, or average over the time window you choose.
What if I only have time to track three metrics?
Signups, MRR and churn rate. They cover acquisition, revenue and retention – the three questions that matter most in the first year.