Explain metric variance and null values
Prevent misleading reports when providers define, backfill, or withhold metrics differently.
Null means the metric is unsupported or unavailable for that resource, while zero means the provider reported no activity. Variance can come from provider definitions, delayed attribution, privacy thresholds, edits, deletions, or different snapshot times.
What this means in practice
Null means the metric is unsupported or unavailable for that resource, while zero means the provider reported no activity. Variance can come from provider definitions, delayed attribution, privacy thresholds, edits, deletions, or different snapshot times.
Use the documented brand and account boundaries consistently across the scheduler, REST API, SDK, MCP tools, SSE consumers, and webhooks. That makes the same social operation explainable to an operator and reproducible by an integration.
Implementation checklist
Label the provider, resource type, and captured_at time for every metric set.
Render unavailable values as an em dash or 'not provided,' not 0.
Explain whether a dashboard shows a snapshot, latest value, or change over time.
Keep raw provider references for support investigation without exposing credentials.
Document transformations used in any derived engagement rate.
Operational details
A provider can revise historical analytics after spam filtering or attribution updates.
Views, impressions, and reach are not interchangeable.
Comparisons are most useful as directional signals with clear methodology.
Common mistakes to avoid
- Do not rank a provider with sparse data as the worst performer by substituting zeros.
- Do not compare account-level followers with post-level reach in one metric column.
- Do not hide measurement changes when a provider API version changes.