Reputation

Mailbox providers (Gmail, Outlook, Yahoo, Apple Mail) decide whether to deliver your mail to inbox, junk, or refuse it entirely based on sender reputation — a continuously-updated score derived from your bounce rate, complaint rate, content patterns, and recipient engagement.

We track two of the load-bearing signals:

• Bounce rate — bounced / sent over the rolling 14-day window
• Complaint rate — complained / sent over the rolling 14-day window

Both are computed from the per-day counter table behind GET /partner/email/stats.

Status thresholds

bounce > 10%  OR complaint > 0.5%  →  PAUSED      (auto-pause)
bounce > 5%   OR complaint > 0.1%  →  AT_RISK     (warning)
else                                →  HEALTHY

These match AWS SES and the broader industry-standard “watch-list” thresholds.

What happens at each threshold

HEALTHY

Default state. No special handling.

AT_RISK

• Email notification to the partner contact
• Banner on the partner panel dashboard
• Recommendation to investigate recent campaigns + clean lists
• No automatic action against your sending — but you should fix the underlying issue immediately

PAUSED

• All /send calls return 403 SENDING_PAUSED
• The reputation status reflects on GET /partner/email/quotas
• Your panel dashboard shows the alert + reason
• An admin reviews + decides whether to:
  • Drop you back to sandbox (so you can test fixes)
  • Re-enable with a stern warning
  • Keep paused pending list-cleaning evidence

To resume sending after PAUSED:

1. Identify the breach source (which campaigns / list segments caused it)
2. Clean the list — remove every address that bounced / complained
3. Reach out to support with your remediation plan
4. Admin can lift the pause via the partner panel

What we don’t track (yet)

• Engagement signals (open rate, reply rate, forward rate) — on roadmap for shared-IP reputation; today these only matter for dedicated-IP customers.
• Spam-filter scores (SpamAssassin etc.) — you control content; our infrastructure handles authentication.
• Domain age / DMARC alignment — hardened automatically through the verification flow.

Per-config-set vs account-wide

By default reputation is computed at the account level. Sends across all config sets contribute to the same rolling window.

If you want to isolate a high-risk experiment from your main reputation, set reputation_tracking_enabled: false on that config set. Bounces + complaints from those sends are still recorded in your event log + suppression list, but excluded from the rolling reputation calculation.

Use sparingly — false is an opt-out from a healthy default.

Inspecting your reputation

curl https://apis.splashifypro.com/api/v1/partner/email/reputation \
  -H "Authorization: Bearer $SPLASHIFY_API_KEY"

Response:

{
  "success": true,
  "window_days": 14,
  "sent": 12500,
  "bounced": 218,
  "complained": 4,
  "bounce_rate": 0.01744,
  "complaint_rate": 0.00032,
  "status": "HEALTHY"
}

/quotas returns the same status alongside daily-quota info — use that endpoint as the single read-once-per-page source on your dashboard.

How to keep reputation healthy

1. Send to opt-in lists only. Buying lists or scraping email addresses is the #1 cause of complaint-rate breaches.
2. Honor unsubscribes within 10 days. It’s a CAN-SPAM / CASL / DPDP requirement AND it’s how you avoid complaint spikes.
3. Verify before sending. Hitting hundreds of stale addresses once spikes your bounce rate hard.
4. Warm up new identities slowly. Mailbox providers throttle first-contact volume — start with low volume to high-engagement recipients, ramp over a week or two.
5. Watch your DMARC reports. Misaligned mail (sent through us but with wrong From-domain) gets quarantined or rejected — that shows up as bounces in our stats.