Gmail vs Outlook: Why Your Cold Email Lands in Spam and How to Fix It

Almost every cold sender eventually sees the same pattern: their messages inbox cleanly with one provider and consistently spam with another. Most of the time, the provider doing the filtering is Outlook, but Gmail can be just as harsh once a domain crosses the wrong line. The reason is that the two filtering systems weight signals very differently, and tuning for one without considering the other quietly hurts your numbers.

How Gmail decides

Gmail's filter is the most heavily machine-learned of the major providers. It uses domain reputation as the dominant signal, with IP reputation contributing far less than it did five years ago. The reputation score is computed continuously from a blend of user actions (marking as spam, archiving without reading, replying, marking as important, moving out of spam) and technical signals (authentication, list hygiene, complaint rates, encryption).

If you want to know what Gmail thinks of you, sign up for Postmaster Tools. The Domain Reputation tab is the single most useful number in cold email. High means Gmail is inboxing you. Medium means you are on the edge. Low means everything is going to spam. The number updates daily and reacts to behaviour over a rolling seven day window.

Postmaster also exposes spam rate (your complaint rate, target below 0.10%), IP reputation, authentication pass rate, encryption rate, and delivery errors. The Feedback Loop ID metric tells you whether recipients are clicking the spam button on your messages, which is the single most damaging user action you can earn.

How Outlook decides

Outlook (covering Outlook.com, Hotmail, Live, and Microsoft 365 mailboxes) uses a more rules-driven blend, with IP reputation weighted heavier than at Gmail and content scanning that is stricter on certain patterns. Outlook's SmartScreen has been deprecated in name but its logic lives on inside the Exchange Online Protection (EOP) stack.

To see what Outlook thinks, register your IPs with Smart Network Data Services (SNDS) and join the Junk Email Reporting Program (JMRP). SNDS shows you per-IP data: filter result (Green, Yellow, Red), complaint rate, trap hits, and message volume. JMRP gives you complaint feedback similar to Gmail's FBL.

Outlook is less forgiving than Gmail on three things: high link density, missing or weak DKIM, and brand-new domains. It is more forgiving on volume ramp, partly because it sees less of it: most cold sending pipes more mail to Gmail than to Outlook.

Why one inboxes and the other does not

The common asymmetric outcomes and their typical causes:

• ✦Gmail inbox, Outlook junk. Usually weak DKIM (1024-bit or missing), aggressive link-to-text ratio, or an IP on a public blocklist that Outlook honours (such as SpamRats or UCEPROTECT) but Gmail does not.
• ✦Outlook inbox, Gmail spam. Usually high user complaint rate at Gmail, poor warmup behaviour (no replies or spam recovery), or a content pattern Gmail's classifier has learned, such as a tracking pixel from a known cold-email vendor.
• ✦Both inboxing then both spam after a week. Volume ramp tripped both providers at once. Pull back to 30% and rebuild.
• ✦Both spam from day one. Authentication failure on at least one of SPF, DKIM, DMARC. Check the headers, do not guess.

Provider-specific tuning

For Gmail, prioritise user-action signals during warmup. Spam-folder rescue and quoted-reply traffic are weighted heaviest. Keep your link-to-text ratio under 1:5, avoid single-image messages, and never use shortener-of-a-shortener URLs. Set up BIMI once your DMARC is at p=quarantine or stricter, it gives you a visible brand indicator in Gmail and contributes a small positive signal.

For Outlook, prioritise IP reputation if you control the IP. Even on shared pools, send through the same vendor consistently so Outlook can build a stable picture. Keep DKIM at 2048-bit, register every sending IP in SNDS, and monitor the Red counts weekly. Outlook also rewards a slow, steady volume profile more than Gmail does, so resist the urge to batch sends.

Seed testing across both

Open and reply rates do not tell you where your mail landed, they tell you what your engaged subset did. To know actual inbox placement, send to seed mailboxes you control across at least Gmail, Outlook.com, Microsoft 365, Yahoo, and one corporate Microsoft 365 tenant. Check folder placement before every campaign and after any authentication change.

The honest version: most cold senders optimise for Gmail and treat Outlook as a bonus, because Gmail volumes dominate. That is fine, until you sell into enterprise and discover that most decision makers read mail in Outlook. Build the muscle for both early.