Why Are My Cold Emails Going to Spam?
Usually because of one of four layers, in this order: authentication (SPF, DKIM, DMARC), list quality (bounces), sending volume and setup (complaints), then copy. Your sending tool's "delivered" rate won't show it, because spam-folder mail counts as delivered. When replies sink but delivery looks healthy, find out where the mail is landing, provider by provider, and how you're being scored, before anyone rewrites a line.
Updated
What does a "delivered" email really tell you?
Only that the other side's server took it. The folder is invisible to your tool: main inbox, Promotions and junk all score as delivered, and a bounce is the one outcome that doesn't. A campaign can show 98% delivered while one provider is burying it (illustrative example: 98% delivered, 40% inboxed).
| Term | What it measures | Where you see it |
|---|---|---|
| Delivery rate | Share of mail the receiving server accepted | Your sending tool |
| Deliverability / inbox placement | Share of mail that reached the inbox | Placement tests by provider |
| Spam rate (Gmail) | Share of inbox mail that users marked as spam | Google Postmaster Tools |
Why did my replies drop when deliverability looks fine?
Because the dashboards you're looking at can't see the spam folder. Your tool counts spam-folder mail as delivered. And Google's Postmaster Tools spam rate only counts mail that reached the inbox and was then marked as spam; mail Gmail already filtered isn't included (Google, Postmaster Tools help).
Inside a single team it tends to go like this: a tired list segment bounces, a handful of people hit the spam button, one provider quietly moves you to junk, its users stop replying, and your overall rate drifts down. For scale, Woodpecker's 2026 page cites a 5.1% average reply rate (2024), while Instantly's 2026 benchmark puts the average on its platform at 3.43%. These are different datasets, so don't read them as a year-over-year decline. When your own rate drops, ask "where is it landing?" before "what's wrong with the copy?"
How can I tell if my cold emails land in spam?
Three tests, roughly an hour, with the copy frozen until you're done:
- Where it lands. Mail the email that's live right now to seed accounts on Gmail, Outlook, Yahoo and a business domain or two, and note the folder at each. A blended number can't show you one provider burying you. Our inbox placement test page walks through this.
- How you're scored. In Google Postmaster Tools, follow two things for your Gmail traffic: the spam-rate line and the authentication results. A line that started climbing weeks before replies fell is your culprit, with a date on it.
- Which list died. Label every reply (interested, neutral, referral, not interested) and group by the list it came from. One list gone silent while the rest keep producing means bad data, not a bad email.
How do I fix cold emails going to spam?
Repair from the foundation up; a crack lower down makes every fix above it look useless:
| Order | Layer | What to do | Line to hold |
|---|---|---|---|
| 1 | Authentication | All three records passing for the domain you send from. Walkthrough: SPF, DKIM and DMARC explained | Google requires all three for bulk senders (~5,000+ messages a day to Gmail), DMARC at least p=none |
| 2 | List quality | Run every address through verification; drop hard bounces and generic inboxes like info@; re-check any list you reuse | Our guideline is to pause above 2% bounces; Google and Microsoft publish no bounce threshold. Verified lists bounce less, so verify before every send |
| 3 | Volume and setup | Separate domains just for cold mail; grow by adding inboxes, not by loading one up | About 30 cold sends per inbox per day (our guideline); spam complaints under 0.3%, aim for under 0.1% (Google, Yahoo) |
| 4 | Copy | Tighter segments (under 50); follow-ups with something fresh; a visible opt-out | Under-50 campaigns averaged 5.8% replies vs 2.1% for 500+ (Woodpecker's report on 16.5M emails) |
Google says bulk senders above a 0.3% spam rate are ineligible for mitigation, and since November 2025 Gmail has been ramping up enforcement with temporary and permanent rejections. By the time rejections show up as bounces, the damage is already done. New domains also need time before they carry volume; see how long to warm up a domain. To size your sending setup, try the Pipe Playbook calculator.
What makes a spam problem worse?
- Cranking up volume to replace lost replies. More mail on a damaged reputation damages it further. (More in the activity trap.)
- Grading subject line tests by opens. An open isn't proof a person read anything; score tests on replies.
- Jumping to a new sending platform with the identical list and pace. The mess relocates with you; the software was never the cause.
See also: Why don't prospects answer unknown numbers, and what can you do about it?
The sending days, sends per inbox, inboxes per rep and touches per contact on this page are our house assumptions, with one line on why we use each.
Sources
Research and vendor pages accessed October 1, 2026.
- Google, Email sender guidelines: https://support.google.com/mail/answer/81126
- Google, Email sender guidelines FAQ: https://support.google.com/a/answer/14229414
- Google, Postmaster Tools help (how spam rate is calculated): https://support.google.com/mail/answer/14668346
- Yahoo Sender Hub, best practices: https://senders.yahooinc.com/best-practices/
- Woodpecker, Cold Email Statistics (2026), including its report on 16.5M emails: https://woodpecker.co/blog/cold-email-statistics/
- Instantly, Cold Email Benchmark Report 2026: https://instantly.ai/cold-email-benchmark-report-2026