Practice API analytics vocabulary: usage dashboards, calls per minute, top consumers, p95 latency by endpoint, API error rate by consumer, and API health metrics.
0 / 5 completed
1 / 5
In an API analytics dashboard, what does 'calls per minute' (CPM) measure?
Calls per minute is a real-time throughput metric. API teams monitor CPM to detect anomalies (sudden spikes may indicate abuse or viral growth), enforce rate limits, and plan for capacity. It is often broken down by endpoint, consumer, or region.
2 / 5
An analytics dashboard shows 'top consumers.' What information does this provide?
Top consumers ranking shows which applications, tenants, or API keys drive the most traffic. This is essential for identifying power users (for business outreach), detecting abuse (one consumer taking a disproportionate share), and prioritizing SDK/support investments.
3 / 5
What does 'p95 latency by endpoint' mean in API monitoring?
p95 (95th percentile) latency is a key SLO metric: it captures the experience of users who are not quite in the 'worst 5%'. Monitoring p95 by endpoint identifies slow endpoints that may be hurting the majority of heavy users, even if average latency looks acceptable.
4 / 5
Why is 'API error rate by consumer' a more useful metric than overall API error rate?
Segmenting error rates by consumer reveals whether a problem is systemic or isolated. If one consumer has 40% error rate while others have 0.1%, the issue is their integration, not the API itself — enabling targeted support rather than broad incident response.
5 / 5
A team says 'this endpoint is the most called.' What strategic decisions does this information drive?
The most-called endpoint is the most critical: performance degradations there affect the most users, SLO breaches there have the biggest business impact, and reliability investments there have the highest ROI. Analytics reveal where to invest engineering effort.