Mailbox providers track user engagement to determine email placement. Key signals include: opens (diminished by Apple MPP), clicks (strongest remaining signal), replies (highly positive), spam complaints (strongest negative), deletion without reading, time spent reading, and marking as not spam. Providers weight recent engagement heavily—inactive subscribers damage deliverability regardless of past engagement.
Email Engagement Signals: What Mailbox Providers Actually Measure
Why Engagement Matters for Deliverability
Mailbox providers learned that authentication and content analysis aren't enough to identify unwanted email. Legitimate senders with perfect SPF/DKIM/DMARC can still send email users don't want.
The solution: measure how recipients interact with email and use that data to predict whether future emails are wanted.
This creates a feedback loop: engaged recipients → inbox placement → more engagement. Disengaged recipients → spam/promotions → less engagement → worse placement.
The Engagement Signals
Opens
What it is: Email opened in a mail client (pixel loads)
Provider visibility: Indirect—pixel requests come to your tracking domain, but providers can infer from other signals
Current reliability: Low since Apple Mail Privacy Protection (2021). Apple prefetches all images, registering false opens for 50-60% of recipients.
Impact on deliverability: Historically high, now diminished. Providers are shifting to other signals.
Practitioner note: Open rate is dying as a useful metric. I still track it for non-Apple users, but click rate is now my primary engagement indicator. Clients who obsess over open rates are optimizing for a metric that no longer reflects reality.
Clicks
What it is: Recipient clicked a link in your email
Provider visibility: High—providers see outbound HTTP requests from their interface
Current reliability: High—click tracking still works reliably
Impact on deliverability: Strong positive signal. Clicks indicate genuine interest.
Clicks are now the most reliable engagement signal. Build emails that encourage clicks, even if just "click to read more."
Replies
What it is: Recipient replied to your email
Provider visibility: Complete—providers see the reply compose and send
Current reliability: Perfect—no privacy technology affects this
Impact on deliverability: Strongest positive signal. Replies indicate a real relationship.
Transactional and support emails that generate replies build reputation. Marketing emails rarely generate replies, which is one reason marketing reputation is harder to maintain.
Spam Complaints
What it is: Recipient clicked "Report Spam" or "This is spam"
Provider visibility: Complete
Impact on deliverability: Strongest negative signal. A single complaint outweighs many positive interactions.
Gmail's 0.3% threshold means 3 complaints per 1000 inbox deliveries triggers increased filtering. Microsoft and Yahoo have similar thresholds.
Complaints happen for three reasons:
- User genuinely doesn't want the email (your problem)
- User doesn't recognize the sender (your branding problem)
- User uses spam button instead of unsubscribe (partially their behavior, partially your UX)
Deletion Without Reading
What it is: Recipient deleted email without opening
Provider visibility: Complete
Impact: Moderate negative. Occasional deletion is normal. Pattern of deletion signals unwanted email.
Moving to Folders
What it is: Recipient moves email to folders, archives, or categories
Provider visibility: Complete
Impact: Complex. Moving to "Important" or archiving = positive. Moving to trash = negative. Moving to folders = neutral.
Time Spent Reading
What it is: How long recipient has email open
Provider visibility: Estimated via client behavior
Impact: Longer reading time is positive, but hard to measure accurately.
Mark as Not Spam
What it is: Recipient moves email from spam to inbox
Provider visibility: Complete
Impact: Strong positive signal that corrects previous filtering decisions.
How Providers Use Engagement Data
Mailbox providers combine engagement signals into reputation models:
| Signal | Weight | Direction |
|---|---|---|
| Spam complaint | Very High | Negative |
| Reply | High | Positive |
| Click | High | Positive |
| Mark not spam | High | Positive |
| Open (non-Apple) | Medium | Positive |
| Delete unopened | Medium | Negative |
| Ignore repeatedly | Medium | Negative |
| Forward | Low | Positive |
Providers evaluate engagement at multiple levels:
- Per-campaign: How did recipients respond to this specific send?
- Per-recipient: Does this user typically engage with this sender?
- Per-sender: What's the overall engagement pattern for this sender?
Practitioner note: The per-recipient level is what kills you with unengaged subscribers. Gmail knows that Bob hasn't opened your last 15 emails. When you send email 16, Gmail assumes Bob still doesn't want it—regardless of how others engaged.
Measuring Your Engagement
Track these metrics consistently:
Click-through rate (CTR): Clicks / Delivered × 100
- Good: > 2%
- Concerning: < 1%
- Dangerous: < 0.5%
Click-to-open rate (CTOR): Clicks / Opens × 100
- Measures quality of content for people who opened
- Less reliable now due to Apple MPP
Complaint rate: Complaints / Delivered to Inbox × 100
- Target: < 0.1%
- Threshold: 0.3%
- Danger: > 0.5%
Unsubscribe rate: Unsubscribes / Delivered × 100
- Healthy: 0.1-0.5% per campaign
- Spiky: Indicates content mismatch
List engagement distribution:
- Active (clicked in 30 days): Should be > 20%
- Engaged (clicked in 90 days): Should be > 40%
- Inactive (no clicks in 180+ days): Should be < 20%
Optimizing for Engagement
Content that drives engagement
- Clear CTAs: Every email should have an obvious next step
- Personalization: Use data you have to make content relevant
- Value delivery: Give something useful, not just asks
- Proper frequency: Too much = fatigue, too little = forgotten
List management for engagement
- Engagement-based segmentation: Send more to engaged, less to unengaged
- Sunset policies: Remove subscribers who haven't engaged in 6-12 months
- Re-engagement campaigns: Attempt recovery before removal
- Preference centers: Let subscribers choose frequency and content type
Technical optimization
- Segment transactional/marketing: Protect transactional reputation
- Send time optimization: Test when your audience engages
- Mobile optimization: 60%+ read on mobile—broken mobile = no engagement
The Engagement Decay Problem
Engagement decays over time:
| Time Since Last Engagement | Typical Deliverability Impact |
|---|---|
| 0-30 days | Full inbox placement |
| 31-60 days | Slight filtering increase |
| 61-90 days | Noticeable filtering |
| 91-180 days | Significant spam risk |
| 180+ days | High spam risk |
This is why list age correlates with deliverability problems. A "great list" from 2023 is a liability in 2026 without ongoing engagement.
Practitioner note: I've audited lists where 70% of subscribers hadn't engaged in over a year. The client was confused why deliverability was terrible—they had "a million subscribers." They had maybe 200,000 engaged subscribers and 800,000 weights dragging them into spam.
When Engagement Can't Save You
Engagement helps inbox placement, but it can't overcome:
- Blocklist presence: Spamhaus listing = blocked regardless of engagement
- Authentication failure: SPF/DKIM/DMARC problems override engagement
- Infrastructure issues: Poor IP reputation, missing PTR records
- Content disasters: Malware links, phishing patterns
Fix technical issues first. Engagement optimization matters only after the foundation is solid.
If your engagement metrics are healthy but deliverability is poor, the problem is likely technical, not behavioral. Schedule a consultation for a systematic diagnosis.
Sources
- Gmail: Email Sender Guidelines
- Return Path: Email Engagement and Deliverability
- Validity: Email Benchmark Reports
- Litmus: State of Email Engagement
v1.0 · March 2026
Frequently Asked Questions
What's the most important email engagement signal?
For negative impact: spam complaints. One complaint hurts more than dozens of opens help. For positive impact: replies are strongest, followed by clicks. Opens matter less since Apple Mail Privacy Protection made them unreliable for 50%+ of recipients.
How do mailbox providers track engagement?
Opens are tracked via pixel loads (unreliable with Apple MPP). Clicks are tracked via redirect links. Replies, spam reports, and deletions are tracked directly by the mailbox provider. You can't see this data, but providers use it to filter your future emails.
What engagement rate do I need for good deliverability?
There's no universal threshold, but industry benchmarks: 20-30% open rate (pre-MPP measurement), 2-5% click rate, <0.3% complaint rate. What matters more is relative engagement—are users engaging more or less than last month? Declining trends trigger filtering.
How long does engagement history affect deliverability?
Mailbox providers weight recent engagement heavily. The last 30-90 days matter most. A subscriber who engaged six months ago but ignored the last 10 emails is now hurting your deliverability, not helping it.
Do unopened emails hurt deliverability?
Yes, but indirectly. Consistently ignored emails signal that recipients don't want your mail. Mailbox providers learn this pattern and start filtering proactively. This is why sunset policies exist—removing unengaged subscribers protects deliverability for engaged ones.
Want this handled for you?
Free 30-minute strategy call. Walk away with a plan either way.