Microsoft SNDS: how to check and fix your IP reputation with Outlook

Microsoft's Smart Network Data Services (SNDS) gives senders direct insight into how Outlook's consumer mail infrastructure sees their sending IPs. The data exists because Microsoft realised that legitimate senders fix problems faster when they have visibility, and a sender who fixes problems faster is one less spam complaint to process.

What SNDS actually shows

SNDS publishes a per-IP, per-day report covering traffic to Microsoft consumer mailboxes (outlook.com, hotmail.com, live.com, msn.com). For every IP you send from, every day, you get:

• RCPT count. How many recipients you attempted to deliver to.
• Data count. How many of those recipients you actually sent message bodies to.
• Message recipients. Total recipient mailboxes touched.
• Filter result. Green / Yellow / Red — Microsoft's internal classification of your IP's reputation that day.
• Complaint rate. Percentage of recipients who hit the "Junk" or "Phishing" button.
• Trap hits. How many of your messages landed in Microsoft's spam-trap mailboxes.
• Sample HELO and FROM data. What you announced in the SMTP conversation.

Critically: this is real production data, refreshed daily, for the actual IP you're using. Nobody else gives you this. Gmail's Postmaster Tools is per-domain and lags more; SNDS is per-IP and same-day.

How to register an IP

Visit sendersupport.olc.protection.outlook.com/snds/ and create an account using a personal email (Microsoft account or any verified email). The account is the registration anchor.

For each IP you want data on, you need to prove control via DNS:

1. On the SNDS dashboard, request access to the IP. SNDS generates a unique token and shows you the TXT record to publish.
2. Add the TXT record under your domain. Microsoft accepts either reverse DNS (rDNS) of the IP or a TXT record at the IP's rDNS hostname. Most senders use the rDNS path.
3. Wait up to 24 hours for DNS propagation, then click "Verify" on the dashboard. SNDS confirms and access begins for that IP.
4. Repeat per IP. You can manage hundreds of IPs in one account.

If you're on a shared ESP IP, you cannot register it yourself — your ESP's deliverability team has access and should share aggregated data with you on request.

Reading the data: colour codes

The most actionable column is the filter result. Microsoft uses a traffic-light system:

• Green. Your IP is in good standing. Mail accepted and largely placed in inbox subject to per-recipient factors.
• Yellow. Moderate concern. Mail is accepted but more likely to be filtered to junk on a per-recipient basis. Trap hits or complaint rate are starting to register.
• Red. Significant problem. Mail likely going to junk for most recipients, may be rejected outright at the gateway depending on which downstream filter decides.

The status reflects the worst signals over the most recent 24-hour window. A single bad day with a trap hit can move you from Green to Yellow; sustained Yellow can move you to Red within a few days.

Fixing a Red status

Red is recoverable but only with deliberate action. The playbook:

1. Pause sending to outlook.com / hotmail.com / live.com immediately. Continued volume on a Red IP makes the status persist longer.
2. Identify the trigger. SNDS shows trap hits and complaint rate. If trap hits are nonzero, you have list- quality problems and need to find the trap addresses. If complaint rate is high (above 0.3%), your content or your audience targeting is off.
3. Audit the most recent imports. Drop anything unverified. Run remaining addresses through a list validator.
4. Sign up for JMRP. Microsoft's Junk Mail Reporting Program is a feedback loop that delivers complaint data to your designated address. You'll see exactly which sends triggered which complaints, with full headers. The complement to SNDS.
5. Resume Microsoft sending at low volume. Start at maybe 5% of pre-Red volume to that domain group. Watch SNDS daily. Status should recover from Red to Yellow within 5–7 days, then to Green over a further 7–14 days if no new bad signals occur.
6. Submit a sender support ticket if status persists. The form at sender.office.com/sender is the right place to escalate Red status that doesn't recover after remediation. Provide SNDS data, JMRP findings, and your remediation steps.

Trap hits — the most important number

SNDS's trap-hit count is the single most important column. Microsoft operates a network of spam-trap mailboxes on outlook.com and adjacent domains. Some are pristine traps (addresses that never belonged to a real user), some are recycled (formerly real addresses that expired and were converted to traps after a long dormancy period).

A pristine trap hit is the worst signal — it means you got an address from somewhere that scrapes or generates addresses, since no human ever opted in. Recycled trap hits indicate stale list quality (you're sending to addresses that haven't been valid for years).

Either way, the fix is list hygiene. Sunset unengaged recipients aggressively. Validate every import. Drop any Microsoft address that has not engaged in the past 6 months.

SNDS vs Office 365 (the business mail world)

Critical distinction: SNDS covers consumer Outlook (outlook.com, hotmail.com, live.com, msn.com). It does not cover Office 365 / Exchange Online business mailboxes — the corporate side, where your customers' @company.com addresses live if they use Microsoft 365.

For Office 365 deliverability, Microsoft offers Smart Network for Senders (separate program) and the Office 365 Sender Score within their broader Microsoft Defender infrastructure. Different filtering pipeline, different signals, different remediation paths.

Practical implication: SNDS gives you visibility on consumer Microsoft mail. If your audience is heavily B2B, SNDS data is helpful but not complete. Use SNDS plus a weekly placement test against Office 365 seed mailboxes to cover both surfaces.

Building daily monitoring

SNDS does not push alerts. You have to pull. The minimum working setup:

• Daily download of the SNDS data file (CSV, available via authenticated URL). Microsoft publishes a stable access URL per account.
• Parse the CSV per IP, alert on any move from Green to Yellow or Red, alert on any trap hit, alert on complaint rate above 0.1%.
• Plot trends over time so you see the slow drift before the colour change.
• Cross-reference SNDS data against your sending log so you can correlate spikes back to specific campaigns.

Common mistakes

• Registering once, never checking. SNDS data only helps if you read it. Build a dashboard or a daily email digest.
• Ignoring Yellow status. Yellow is the warning before Red. Investigate Yellow as if it were Red, fix while it's still cheap.
• Continuing volume during recovery. Sending heavy volume on a Yellow or Red IP extends the recovery window. Pause, fix, ramp.
• Not pairing SNDS with JMRP. SNDS is incomplete without complaint feedback. Sign up for both.
• Treating SNDS as Office 365 data. It's consumer-only. Don't make Office 365 decisions based purely on SNDS.

Frequently asked questions