I would like to suggest stronger pre-delivery filtering for messages that already exhibit multiple independent signs of coordinated spam or spoofing.

I am not suggesting that every message with one failed SPF, DKIM, or DMARC result should automatically be rejected. Legitimate forwarding, mailing lists, and configuration mistakes can sometimes interfere with authentication.

Instead, I suggest using a layered pre-delivery rejection model in which several high-risk signals together can trigger rejection at the SMTP edge before the message is accepted into a user's mailbox at all.

Examples of signals that could be combined:

• SPF hard failure, DKIM failure/absence, or DMARC failure
• A sending IP or small IP range suddenly producing many similar messages
• Rapid rotation of unrelated visible From domains
• Nearly identical HTML layouts, wording, image placement, subjects, or campaign structure
• Multiple suspicious redirect or short-link URLs
• A sender/domain with no prior relationship to the recipient
• Poor or newly deteriorating IP/domain reputation
• Repeated user spam complaints tied to the same campaign fingerprint

A recent spam burst I encountered illustrates the problem. Multiple messages arrived within minutes of each other from the same small network range. The visible sender domains changed repeatedly, authentication frequently failed or was incomplete, and the messages used nearly identical photo/romance-style templates and redirect links.

The system correctly placed the messages in Spam, but by that point they had already been accepted into the mailbox.

My suggestion is:

1. Use campaign-level fingerprinting across messages, not just sender-address blocking.

2. When multiple strong indicators agree, reject the message during SMTP rather than depositing it into the Spam folder.

3. For uncertain cases, use temporary SMTP deferral/greylisting while additional reputation and URL checks are performed.

4. Safely inspect and expand redirect URLs before delivery, including URLs hosted on otherwise reputable services that may be abused as redirectors.

5. Escalate filtering automatically when a narrow IP range begins sending repeated, structurally similar spam using rotating domains.

6. Preserve the Spam folder primarily for genuinely uncertain messages rather than messages already classified with very high confidence as abusive.

The goal would not be to make SPF/DKIM/DMARC individually mandatory for every legitimate message. The goal would be to combine authentication, reputation, campaign fingerprints, URL analysis, and recipient history into a stronger pre-delivery decision.

If several independent signals all point to the same coordinated spam campaign, I believe the receiving provider should increasingly reject subsequent copies before they ever appear in users' Spam folders.

Thank you for considering stronger campaign-level and pre-delivery spam enforcement.