How to automate alerts when placement suddenly shifts?

Sudden placement shifts cost money every hour you don't catch them. The fix is a monitoring stack that pings a human within minutes when inbox rate drops, spam complaints spike, or bounces jump above your baseline. Manual dashboard checks once a day are not enough. By the time you notice, you have already sent a million emails into the spam folder.

Here is the stack we recommend for a working alerting setup.

Layer 1: Seed-list inbox placement. Tools like GlockApps, Validity Everest, or Inboxally drop test emails into seed accounts across Gmail, Yahoo, Outlook, Apple, and regional providers, then report where each one landed. Run a seed test on every campaign and set alerts when Gmail inbox placement drops more than 10 points from your rolling 7-day average. Seed tests are not perfect (the seed accounts behave differently than real subscribers), but they catch big shifts fast. For why Gmail can drop while Yahoo holds steady, see why different MBPs produce different inbox results.

Layer 2: Google Postmaster Tools and SNDS. Google publishes domain reputation, IP reputation, spam rate, and authentication rates for senders pushing more than a few thousand messages per day to Gmail. Microsoft SNDS does the equivalent for Outlook IPs. Neither has native alerting, so pipe the data out with the Postmaster Tools API into a Google Sheet or a small Node script, then alert when spam rate crosses 0.10% (Google's soft warning) or 0.30% (Google's hard threshold). Gmail's bulk sender guidelines make the 0.30% line non-negotiable.

Layer 3: ESP webhooks. Your ESP (SendGrid, Mailgun, Postmark, Customer.io, Klaviyo) emits webhooks for bounces, complaints, and unsubscribes in real time. Forward those into a tool that can do threshold math. We use a simple n8n flow that aggregates events in a 15-minute window and posts to Slack when:

• Hard bounce rate above 2% in a 15-minute window
• Complaint rate above 0.1% in a 15-minute window
• Soft bounce rate above 5% sustained for an hour
• Any single campaign with more than 20 Gmail complaints in an hour

Layer 4: Blocklist monitoring. Subscribe to MXToolbox blacklist monitoring or run your own checks against Spamhaus, SURBL, and SORBS every 15 minutes. A fresh listing on Spamhaus SBL or XBL will tank your delivery within an hour, and you need to know before your sending IP gets fully blackholed.

Routing the alerts. Slack is fine for the first ping, but you need an on-call rotation for the bad ones. PagerDuty, Opsgenie, or Better Stack handle the escalation: Slack first, SMS at 10 minutes if nobody acknowledges, phone call at 20 minutes. Tier the rules so a 0.12% spam rate posts to a channel and goes to sleep, but a Spamhaus listing or a 1% complaint rate wakes someone up.

Picking thresholds without flooding yourself. Watch your numbers for two weeks before turning alerts on. If your Gmail inbox rate normally bounces between 86% and 92%, do not alert at 90% or you will hate your life by Wednesday. Alert at 80%. If your complaint rate sits around 0.04%, alert at 0.10%, not 0.05%. The goal is to catch the drops that mean something, not every tick of normal variance. Read more on what counts as normal variance in how often MBPs update filtering rules.

Write the runbook before you need it. When the alert fires at 2am, the on-call person should not be googling "what is a Spamhaus listing." Have a one-page doc per alert type: what it means, first three things to check, who to escalate to, how to pause sending. We keep ours in Notion linked directly from the Slack alert message.