Reliability
How We Handle High-Volume DM Sending Without Breaking Meta's Rules
August 29, 2026 · 6 min read
When a campaign takes off and a post gets hundreds of comments in a short window, sending every DM instantly and all at once isn't actually the safest or most reliable approach. Here's how social.fluxx handles high-volume sending in a way that stays reliable and stays within Meta's own rate limits.
Every send goes through a queue, not straight out
Rather than firing off a DM the instant a matching comment is detected, each send is placed into a queue and processed by a dedicated background worker. This decouples the moment a comment comes in from the moment the DM actually goes out, which is what makes pacing possible in the first place.
Rate limiting is enforced centrally, per account
The queue enforces Meta's documented 750-per-hour private reply cap per connected Instagram account, pacing sends so a busy campaign never exceeds it. This limiting is applied centrally across the whole system rather than per individual process, which matters because it's the account's limit that counts, not any one worker's activity.
Retries and reliability
If a send fails, whether from a transient network issue or a temporary API error, the worker retries it with backoff rather than immediately hammering the API again, which is both more reliable and gentler on Meta's systems. Every send outcome, sent, queued, or failed, is recorded in the DM log, so you can see exactly what happened to any specific message.
The takeaway
A queue and a rate-limiting worker sitting between a comment and a DM isn't extra overhead, it's what makes a busy campaign reliable and keeps the connected account within Meta's own limits.
Start free