Every DM, on the record.
We publish a live snapshot of what Replio does so creators can see — in numbers — the safeguards behind every reply. Updated every hour.
Transparency report, October 2026. Generated Oct 26, 2026, 8:52 PM · updates hourly
What we do differently
Ban Guard
Per-account rate limits, randomised jitter, A/B variants by default, and 3× retries with exponential backoff. Everything logged to a persistent queue — nothing runs in memory.
Scheduled pauses
Users pick quiet hours, quiet days, and vacation windows in their timezone. DMs that would land inside are rescheduled, not skipped — the reply arrives when the day begins.
Dry Run
Before activating a rule, paste real comments and see exactly who would match, which variant would be picked, and how quiet hours would shift the send time. Zero DMs sent during the test.
Explain this DM
Every event shows a full audit trail: matched keyword, selected variant and why, applied delay, queue position, attempt count. The inbox is no longer a black box.
Data hygiene
Instagram access tokens are encrypted with libsodium. You can delete your account in Settings. See our Privacy Policy for details of data processing and retention.
How we compute these numbers
Transparency has to be auditable. Here's exactly what each cell counts.
- DMs sent: rule_events.status = 'sent' this calendar month (UTC).
- Success rate: 100 × sent / (sent + failed) — excludes skipped events such as duplicate commenters or paused rules.
- Held for quiet hours: pending_dms rows flagged quiet_hours_deferred = true this month.
- Average applied delay: mean of applied_delay_ms across queued sends — includes rate-limit jitter and quiet-hours shifts.
- Active creators: connected Instagram accounts whose token is not disabled. Disabled = Meta revoked or user disconnected.
Questions, audits, or press queries? Email replio@jocage.com.