KPI tracking
KPI tracking: turn metrics into actionable alerts
Choose useful KPI indicators, define clear thresholds, and connect KPI tracking with reporting and scheduled Slack alerts.
Give the KPI a next step
Illustration · scheduled checks every 12 or 24 hours
3. Queue a Slack notificationKPI tracking works best when a number leads to a decision. A report can show that revenue missed its target. An alert can prompt the team to investigate before the next review. Kisu adds that scheduled check to metrics you already keep in BigQuery or Google Sheets.
KPI metrics, indicators, and targets: what is the difference?
A metric is a measurement, such as weekly orders. A key performance indicator (KPI) is a metric tied to an important business goal. A target defines the result you want. People sometimes call these “KPI metrics” or “KPI indicators.” The useful distinction is not the label. It is whether the number has an owner and a clear action.
For example, website visits are a metric. Qualified leads per week can be a KPI for a team that needs more sales opportunities. “Review acquisition channels if qualified leads fall below our agreed floor” turns that KPI into an alert condition.
Build a small KPI tracking plan
Start with three to five measures. For each one, record the definition, time window, data source, target, owner, and response. These example rules are starting points, not industry benchmarks:
| KPI | Review condition | Next action |
|---|---|---|
| Qualified leads, last 7 days | Fewer than 40 | Check campaign delivery and lead routing |
| Completed orders, last 24 hours | Fewer than 100 | Check checkout and order imports |
| Open pipeline for this quarter | Below your coverage floor | Review deal quality and creation pace |
Define “qualified,” “completed,” and “open” before you automate. Use one currency and an agreed timezone. Otherwise, the dashboard, report, and alert can all give different answers to the same question.
Keep KPI reporting; add a check between reports
KPI reporting explains trends and trade-offs. Alerts answer a narrower question: does this condition need attention now? Keep your weekly report for context. Use Kisu for a few conditions that justify a message to the team.
In BigQuery, write a single SELECT that returns one row and one Boolean column. In Google Sheets, select a numeric column and choose Sum, Average, Minimum, or Maximum, then a comparison and threshold. Kisu is an alerting layer, not a replacement for your reporting tool.
Set up your first scheduled KPI alert
- Sign in, activate a subscription, and connect BigQuery or Google Sheets in Configuration.
- Add a Slack destination and send a test message. See the Slack webhook guide if needed.
- Create an alert named after the decision, such as “Weekly qualified leads below review floor.”
- Set the condition and choose checks every 12 or 24 hours. For BigQuery, test the query before saving.
- Review History to distinguish the check result from the notification delivery status.
Avoid turning every metric into a notification
Kisu queues a notification on every successful matching check. A condition that stays true can send repeated messages. Pick a threshold that warrants that cadence, and assign someone to investigate. No-data results and failed checks do not mean the KPI is healthy; review them in History.
Start with one KPI that someone currently checks by hand. If that check reliably leads to a useful action, add the next one. For a concrete example, read about marketing KPIs to monitor between reports.