How does Microsoft filter email?

Still have a question, spotted an error, or have a better explanation or a source we should cite?

If a chunk of your list uses Outlook or Microsoft 365, understanding how Microsoft's filters work isn't optional. It's one of the most opinionated filtering systems out there, and it behaves differently from Gmail in ways that trip up even experienced senders.

Microsoft filtering runs in layers. Here's how it actually works from the moment your email arrives.

Layer 1: Authentication and reputation
The first check is whether you've set up SPF, DKIM, and DMARC correctly. Fail those, and you're in trouble before Microsoft even looks at your content. Alongside authentication, Microsoft evaluates your sending IP and domain reputation. Known bad actors get blocked immediately. This is non-negotiable.

Layer 2: Content scanning via SmartScreen
SmartScreen is Microsoft's proprietary content filter. It scans message text, links, attachments, and sending patterns. It's trained on signal from millions of mailboxes, so it's not just keyword matching. Spammy phrases, suspicious links, and unusual sending volumes all feed into its score.

Layer 3: User engagement signals
This is where Microsoft diverges from purely technical filtering. On consumer accounts (Outlook.com, Hotmail), Microsoft watches what recipients actually do with your emails. Opens and clicks are positive signals. Deleting without reading, marking as junk, or letting messages sit unread are negative ones. Enough negative signals and your future messages start heading to Junk for that recipient, and eventually it can affect your reputation more broadly.

Layer 4: Organizational policy (business accounts)
For Microsoft 365 business accounts, there's an additional layer. Exchange Online Protection (EOP) sits in front of every business mailbox, and IT administrators can add their own filtering rules on top of Microsoft's defaults. So even a sender with perfect reputation might get blocked if a company's IT team has added custom restrictions. That's outside your control, but worth knowing.

What this means practically: Microsoft rewards senders whose recipients engage and punishes those who send to people who've lost interest. Keeping your list clean, removing unengaged subscribers before they start ignoring you, and authenticating your domain properly are the three biggest things you can do to stay on Microsoft's good side. (Everything else is fine-tuning.)

If you want to see exactly how Microsoft sees your sending IP or domain, their SNDS and JMRP tools give you a direct window into that data. Worth checking if your Outlook deliverability ever starts slipping.

Contributors

Who worked on this answer

Every name links to their profile. Every company links to their site. Real people, real accountability.

Ask an AI · tailored to your setup

Paste in your domain, sending IP, and any bounce or deferral messages you're seeing, AI will help you narrow down which layer is the problem.

We send email to a list that includes Outlook.com and Microsoft 365 users. Based on the four filtering layers Microsoft uses (authentication, SmartScreen, engagement signals, and Exchange Online Protection), help me identify which layer is most likely causing our deliverability issues. For each layer, give me: (1) the most common failure mode, (2) how to diagnose it, and (3) the fix ranked by impact.

Edit the yellow boxes, then send to the AI of your choice.