Microsoft SNDS for Cold Email: Setup, Filter Colors, and Why Most Senders Cannot Use It

Every cold email team eventually asks the same question about Outlook that they asked about Gmail: what does the receiving side actually think of us? For Gmail there is a clean answer, and it is Google Postmaster Tools. For Microsoft the answer is SNDS, and the guides that recommend it almost never mention that most of the people reading them are ineligible to use it.

This piece covers what SNDS reports, what the thresholds mean, how to enroll if you qualify, and what to do instead if you do not. The second half is the part nobody writes.

What is Microsoft SNDS?

Smart Network Data Services is a free Microsoft postmaster tool that shows how mail from a given IP address performs against Microsoft's consumer mail systems. For each IP and each day, it reports the volume of messages Microsoft accepted, a filter result of Green, Yellow, or Red, the complaint rate from users who marked your mail as junk, the number of spam trap addresses you hit, and sample HELO strings the sending server advertised.

It is the closest thing Microsoft offers to Google Postmaster Tools, and the two are built on different keys. Google reports on your sending domain. Microsoft reports on your sending IP. That difference sounds academic and turns out to decide who can use the tool at all.

The portal has moved. The old address, sendersupport.olc.protection.outlook.com/snds, now returns a permanent redirect to substrate.office.com/ip-domain-management-snds/snds. Bookmarks and older tutorials still point at the legacy URL, which works, but you will land somewhere that looks different from the screenshots.

Why most cold email senders cannot use SNDS

To enroll, you request access for specific IP addresses, an IP range, or an ASN, and then complete an authorization step proving you are responsible for those IPs. Microsoft restricts the data this way for a sound privacy reason: IP-level reputation would otherwise expose one tenant's sending behavior to anyone who guessed the address.

Now think about how a cold email team actually sends. The standard setup is a set of Google Workspace or Microsoft 365 mailboxes, each on a separate domain, sending thirty to fifty messages a day. Those messages leave Google's or Microsoft's outbound IPs, shared with millions of other tenants. You do not own them, you cannot authorize them, and you would learn nothing from the reputation of an IP pool whose behavior is dominated by strangers. The same applies to a shared-IP ESP. If you cannot answer the question "which IP is this mail leaving from, and do I control it?", SNDS has nothing to offer you.

There is a second disqualifier stacked on top of the first, and it is the one that catches teams who do own their IPs. SNDS reports on Microsoft's consumer mail: Outlook.com, Hotmail.com, Live.com. It does not report on Microsoft 365 or Exchange Online business mailboxes. If you sell B2B, essentially all of your Microsoft-hosted prospects sit behind a corporate Microsoft 365 tenant, which means the very recipients you care about are outside the dataset. A team sending to consumers on Hotmail gets real value here. A team selling software to companies gets a dashboard about an audience it is not emailing.

The honest summary is this. SNDS is useful if you run your own mail transfer agent, or you bought a dedicated IP on Amazon SES or a similar provider, and a meaningful share of your recipients use consumer Microsoft mailboxes. That is a narrow intersection, and if you are reading a cold email blog you are probably outside it. Our guide to shared IPs versus dedicated IPs covers when owning the IP is worth the trouble in the first place.

How do I sign up for Microsoft SNDS?

If you qualify, enrollment takes about ten minutes and then a day or two of waiting.

1. Sign in at the SNDS portal with a Microsoft account. A personal Outlook.com account is fine.
2. Choose Request Access and enter the IP addresses, the CIDR range, or the ASN you send from.
3. Complete the authorization step. Microsoft emails a confirmation to an address associated with the IP registration, typically the abuse or postmaster contact in the WHOIS record for that IP block. If you lease the IP from a host, they receive it, and you may need to ask them to forward it.
4. Wait. Data appears within roughly 24 to 48 hours of approval, and it backfills as you send.

One detail catches people out on day one and reads as a broken account: SNDS shows no data for an IP on any day it sent fewer than 100 messages. Cold email volume per IP is often below that line, so a new enrollment can look empty for weeks. This is the same volume-threshold trap that makes Google Postmaster Tools look broken for small senders, and it has the same root cause. Reporting on tiny volumes would let anyone probe an individual's mail behavior.

What do the SNDS filter result colors mean?

The filter result is the single number most people log in for, and its definition is more brutal than most people assume. It is the share of that IP's mail that Microsoft's content filter classified as spam.

Two of those rows deserve a caveat that vendor guides drop. The 0.3 percent complaint figure and the 10 percent RCPT figure are descriptive statistics Microsoft publishes about the population of senders, phrased as: more than 30 percent of IPs keep their complaint rate under 0.3 percent. They are not published block thresholds. Microsoft does not tell you the number at which it starts filtering, and anyone who quotes one is guessing. What the statistics do tell you is where the healthy third of the internet sits, which is a perfectly good target.

The trap row is the one to take literally. Microsoft's trap accounts solicit no mail whatsoever, so a hit is proof that an address entered your list through scraping, purchase, or a bad append. Our guide to what a spam trap is covers how they get onto clean-looking lists, and cleaning an email list covers getting them off.

What is the difference between SNDS and JMRP?

They answer different questions and you want both. SNDS is aggregate telemetry: how does Microsoft see this IP, in colors and rates. The Junk Mail Reporting Program is a complaint feedback loop: when an Outlook.com user clicks the junk button on your message, JMRP forwards you a copy so you know exactly which recipient complained about exactly which campaign.

SNDS tells you the complaint rate went up. JMRP tells you who complained. Without the second, a rising complaint rate is a mystery you cannot act on beyond guessing at the campaign calendar.

Both are now managed from the same Microsoft account and the same SNDS portal. You authorize your IPs in SNDS, then create the JMRP feed from inside that authorized account, pointing it at a mailbox you monitor. Complaints arrive as individual messages in a standard machine-readable complaint format, which is convenient and immediately annoying: nobody wants to read them one at a time. The teams who get value from JMRP pipe that mailbox somewhere and turn the raw complaint notices into structured rows they can suppress against automatically. A complaint that sits unread in a monitored inbox is a complaint you will receive again.

Does Microsoft have bulk sender requirements like Gmail?

Yes, and they landed in 2025. Microsoft announced authentication requirements for high-volume senders to its consumer domains: any domain sending more than 5,000 messages a day to Outlook.com, Hotmail.com, or Live.com must publish valid SPF, valid DKIM, and a DMARC record at a minimum policy of p=none, aligned to either SPF or DKIM. Microsoft also asks for a valid, replyable From and Reply-To address and functional one-click unsubscribe.

Enforcement started on 5 May 2025 by routing non-compliant high-volume mail to the Junk folder, with Microsoft stating it would escalate later to outright rejection, returning the SMTP error 550 5.7.515 Access denied, sending domain [domain] does not meet the required authentication level.

Here is the part that matters for cold email specifically. The 5,000-per-day threshold is measured against consumer Microsoft domains, and a properly run cold email operation sends thirty to fifty messages per inbox per day across a set of domains, of which only a fraction reach Hotmail or Outlook.com addresses. Almost nobody doing real B2B outbound trips the threshold. That is not a reason to ignore the requirements. SPF, DKIM, and DMARC have been table stakes for every mailbox provider for years, and a domain without them is filtered long before volume enters the conversation. Our SPF, DKIM, and DMARC setup guide walks the records line by line.

What should I use if I cannot use SNDS?

Most readers of this page are in this bucket, so treat it as the main section rather than the consolation prize. Sending from Google Workspace or Microsoft 365 mailboxes means you have no IP to enroll, and you need a different instrument panel.

Watch bounce codes, because they are the receiving server talking to you directly. Microsoft's rejection messages are unusually specific, and a 550 5.7.x series response names the reason. Track them per sending domain rather than in aggregate, and read our breakdown of soft bounces versus hard bounces before you decide what to retry.

Run seed-list inbox placement tests. A test that plants addresses across Outlook.com, Microsoft 365, and Gmail and reports where each copy landed gives you the placement answer SNDS would have given you, for the audience you actually email, without needing to own an IP. It is a sample rather than a census, and it is the best available substitute.

Use Google Postmaster Tools for the Gmail half, because it is domain-based and every sender qualifies. Between Postmaster Tools for Gmail and seed testing for Microsoft, you have coverage of the two providers that account for most B2B inboxes.

And watch reply rate per domain, which is the metric no proxy, filter, or reporting threshold can distort. When a domain's replies fall while another domain's hold steady, the first domain has a placement problem, and you will usually see it before any dashboard does. If the numbers stop making sense together, run a full deliverability audit rather than chasing one chart.

Is SNDS worth setting up?

If you own your sending IPs and you email consumers, yes, immediately, along with JMRP. It is free, it is the only direct read on Microsoft's opinion of your infrastructure, and the complaint feed is the fastest way to find the campaign that is quietly generating junk-button clicks.

If you send B2B cold email from Google Workspace or Microsoft 365 mailboxes, which is the overwhelming majority of readers, SNDS is not available to you and would not describe your recipients if it were. Spending an afternoon trying to authorize IPs you do not own is time better spent on the things that actually move Microsoft placement: authentication records, verified lists, per-inbox volume caps, warmed domains, and a message people do not report. Those are the same things that move Gmail placement, which is convenient, because you were going to do them anyway.

For the wider picture, our guide to improving email deliverability covers the levers in order of impact, and cold email deliverability explains how ColdMailer handles authentication checks, warmup, and per-inbox volume control when you send over your own SMTP.