Six reasons your cold email isn't working

Diagnose in this order, because each cause invalidates every test below it. If your mail is in the spam folder, nothing you learn about subject lines is real.

1. It is in the spam folder and you do not know

This is the most common cause and the hardest to see, because every sequencer reports "delivered" when the receiving server returned a 250. That is an SMTP acceptance, not a folder. Spam-folder delivery and inbox delivery look identical from the sender's side.

The tell is a campaign with no replies and no bounces. Bounces mean your mail is being processed; silence with clean delivery usually means filtering.

Check, in this order:

• Authentication, from the recipient side. Send to a mailbox you control, open the raw headers, and confirm spf=pass, dkim=pass with your own domain in the DKIM d= tag, and dmarc=pass. Alignment matters as much as passing — a DKIM signature on a vendor's domain means your reputation is accruing to them.
• Inbox placement per provider. Gmail, Outlook.com, Yahoo and corporate Microsoft 365 tenants all behave differently, and an average hides the one that is failing. Note also that Gmail's Promotions tab counts as delivered in most tooling and reads as a clean hit while replies quietly do not arrive.
• Blocklist status for your sending domain and IP.
• Warm-up. A mailbox with no sending history gets filtered regardless of what it sends.

Fix this first. Everything else below is unmeasurable until you have.

2. The list is rotten

A bounce rate that is climbing is the earliest warning signal you get, and it compounds: bounces damage reputation, which worsens placement, which reduces engagement, which worsens reputation.

Verify addresses before sending. Drop role accounts — info@, sales@, security@ — which bounce more, convert less and are often staffed by someone whose job includes blocking senders. Drop catch-all domains you cannot confirm. And if the list was purchased or scraped wholesale, the correct action is to stop using it: it is a bounce problem, a complaint problem, and under GDPR a lawful-basis problem you cannot document.

3. The segment is too wide

If you cannot write one sentence that is true of everyone on the list, the email will be generic, and generic is what gets ignored. "B2B SaaS companies that just posted their first SDR role" is a segment. "SaaS decision makers" is a list.

A tight segment fixes three things at once: the opening line becomes obvious, the ask becomes specific, and the list becomes small enough to verify properly. Narrowing is almost always a bigger lever than rewriting.

4. The email is too long

Three to five sentences, one ask, one link at most, no image, no attachment. Longer than that and the reader triages it as marketing on a phone screen before reading a word.

Length also reads structurally. A designed HTML template with a hero image and a button contradicts a message that claims to be personal, and recipients notice the contradiction faster than any filter does. For cold outreach, send plain text.

5. The follow-up pattern is generating complaints instead of replies

Three messages: an opener and two follow-ups, three to five business days apart, each shorter and each adding one new thing. Then stop.

A fourth and fifth bump buys very few replies and a meaningful number of complaints, and complaints are the metric that ends your sending — Google asks bulk senders to keep reported spam below 0.3%. "Just bumping this to the top of your inbox" adds nothing and announces that you have nothing to add.

Also check your stop conditions actually work. Test by hand: a reply, an out-of-office, a reply from a colleague, a hard bounce, an unsubscribe. A follow-up sent to someone who already declined is the most expensive message in the sequence.

6. The offer asks for too much

The ask in a first cold email should be small enough that a stranger can grant it without a meeting: a one-page summary, a yes or no, a two-week pilot on one workspace. A calendar link in message one asks for commitment before you have earned attention, and it adds a third-party domain to the email.

Give the reader an easy no, too. "Is this the wrong quarter?" produces one-word replies, and a no is a good outcome: it removes someone from your list before they remove themselves with the spam button.

What not to do while diagnosing

Do not change five things at once. Provider, list source, copy and volume in the same week teaches you nothing, and you will repeat it in two months.

Do not raise volume to compensate. Volume multiplies whatever the campaign already is.

Do not optimise subject lines first. It is the most satisfying lever and the least likely to be the cause.

Do not trust open rate. Apple's Mail Privacy Protection pre-fetches images for Apple Mail users and corporate proxies do the same, so recorded opens include machines. A variant that "wins" on opens may have reached a more Apple-heavy audience.

Do not run a spam-word checker and call it a diagnosis. Filtering is driven by authentication, sending history, reputation, list quality, engagement and complaints. The word "free" is not the problem.

Do not add mailboxes. More senders push the same message to more people. If the message is the problem, you have multiplied it.

Where MailPilot fits

The first item on this list is the one you cannot check from your sequencer. MailPilot measures real inbox placement per provider through a seed-inbox network — including whether Gmail is filing you under Promotions rather than the inbox — warms each mailbox over OAuth on Gmail and Outlook or over SMTP/IMAP for SendGrid, Amazon SES, Mailgun, Postmark and custom SMTP, checks SPF, DKIM and DMARC on every sending domain, and tracks per-mailbox reputation daily. Rule out placement, and the rest of the diagnosis starts producing real answers.