Should You Put Links in Cold Emails? No. Never. Here's Why

Never put a link in a cold email. No URLs, no email addresses, no domain names anywhere in the copy. Two reasons: spam filters score links from unknown senders harshly, and a link converts a reply-seeking email into a click-seeking one, which kills the conversation before it starts.

This is one of our hardest rules, enforced by an automated scan that flags any URL, email address, or domain-like pattern in a cold step before copy ships. It is also the rule clients push back on most, because it feels wrong to talk about a case study without linking to it. It is not wrong. It is the whole trick.

Do links in cold emails hurt deliverability?

Yes, badly. A URL from a sender the recipient's mail server has never seen is one of the oldest spam patterns in existence, because it is exactly what phishing and bulk spam look like: stranger arrives, stranger wants a click. You have no sending history with that domain, no engagement signal, no reputation. The filter has to judge you on the shape of the email alone, and a link is the wrong shape.

It compounds. Links usually mean HTML, and stray HTML in what should be a plain-text one-to-one email is its own flag on our scan. Tracking makes it worse again, which is why we send cold campaigns as plain text with open tracking switched off. Tracked emails carry hidden pixels and rewritten links through a tracking domain, so even an innocent-looking email is silently carrying exactly the machinery filters are built to catch. Our full stance on infrastructure lives in the email deliverability playbook and in our 2026 deliverability guide.

Why do links kill replies even when the email lands?

Because a cold email only has one job: earn a reply. A reply starts a conversation with a human who can be sold to. A click is a dead end. The prospect visits your site, skims for eight seconds, forms a verdict alone, and you never hear about it. You traded a two-way conversation for anonymous traffic.

The reply matters beyond the conversation too. When a recipient replies, the mailbox provider records the strongest positive engagement signal an email can get, which protects every future send from that sender. A click on a link teaches the mailbox provider nothing in your favour. Reply-seeking copy builds sender reputation while it builds pipeline. Click-seeking copy spends reputation to fill a website analytics dashboard.

And structurally, a link competes with your CTA. If the email says "would that be useful?" and also offers a link, you have given the reader two doors, and the low-effort door wins. One email, one ask, and the ask is a reply. We cover what that ask should be in our guide to cold email CTAs and the offer ladder.

What do you do instead of linking?

Describe the asset and let them ask for it. The copy names the specific, tangible thing, a breakdown, a playbook, a sample built for their company, and the CTA is the request: "reply and I'll send it over" or "want me to send it across?" The curiosity a link would have discharged instantly now has exactly one outlet, and it is the outlet you wanted.

This also enables staged delivery, which protects the client's hours. The asset described in the email gets produced or personalised for the prospects who actually reply, not speculatively for ten thousand strangers. The email promises, the reply triggers delivery, and the link finally appears inside a live thread where it belongs.

Same logic for identity. If the copy needs to reference who you are, write the company name as a name, never as a domain. "We're Pipeline Playbooks" reads as a sentence. A bare domain in the body reads as a URL to a filter, because it is one.

What about unsubscribe links, signatures, and the other objections?

Every time we state this rule, the same three objections arrive. They all have answers.

"But we're required to include an unsubscribe link." You are required to give recipients a working way to opt out, and a reply is a working way. A short plain-text line inviting them to reply if they would rather not hear from you does the job without putting a URL in the email. What actually matters is the operational side: every opt-out gets honoured immediately and permanently, which is why suppression lists exist. A prospect who replies "not interested" and never hears from you again is a compliance system working. A footer link nobody trusts enough to click is not.

"But my signature has always had our website in it." A link in the signature is still a link. The filter does not care that it sits below your name. Our sign-off is the sender's first name and the company name on the next line, nothing else. If the prospect wants to look you up, they will type your company name into Google in four seconds, and that search is a stronger signal of interest than any click you could have engineered.

"So links are just banned forever?" No. The ban is on cold steps, because the problem is the combination of an unknown sender and a link. Once the prospect replies, both halves of the problem disappear. The thread is established, the mailbox provider has watched the recipient engage, and a human being is expecting your next message. At that point links are not just acceptable, they are how you deliver on the promise the cold email made. In our system, reply follow-ups are the one place HTML is standard, specifically so the calendar link is a clickable anchor when it is finally natural to send one.

The rule survives because it keeps getting re-tested

We learned to re-run the scan after every edit pass, the hard way. Clients review copy and hand it back with their homepage link helpfully added, or a case study URL dropped into the middle line, and early on those edits could slip through to launch. Now the scan runs again after every single revision, client edits included, and a client asking for a link does not override it. We strip the link, explain why, and describe the asset instead.

The position is absolute because the trade is absurd. What a link buys you is a marginally shorter path for a prospect who was going to reply anyway. What it costs you is inbox placement across the entire campaign and the conversation itself. Describe the thing. Ask if they want it. Send the link when a human asks for it. That order wins.

FAQ

Should you put links in cold emails?

No. Never put a URL, email address, or domain name in a cold email. Links from unknown senders are a classic spam pattern that hurts inbox placement, and a link turns a reply-seeking email into a click-seeking one, killing the conversation before it starts.

Do links in cold emails trigger spam filters?

Yes. A URL from a sender with no history at the recipient's domain is one of the strongest spam signals there is. Our copy scan flags any URL, email address, or domain-like pattern in a cold step, and we send plain text with tracking off.

How do you share content without a link?

Describe the asset in the copy and let them ask for it: "reply and I'll send it over." Once they reply, a human conversation and an engagement signal both exist, and the link lands safely inside an established thread.

Do cold emails need an unsubscribe link?

You need a working opt-out, not a link. A plain-text reply-based opt-out, a line inviting them to reply if they'd rather not hear from you, is a functioning mechanism and keeps the email free of URLs. Honour every opt-out immediately and permanently.

When is it okay to send a calendar link?

Only after the prospect has replied. Once a thread exists, follow-ups can be HTML with a clickable calendar anchor, because the mailbox provider has seen the recipient engage. In the first cold touch, a calendar link is both a spam signal and a premature ask.