Links in email: wrapping, clipping and the domain that gets you filtered

Email is the channel where shortening is most often done for the wrong reason and skipped where it matters.

The wrong reason: shortening every link in a campaign to make the HTML tidy. Your sending platform rewrites all of them through its own click-tracking domain anyway, so you have added a hop, a point of failure and a second set of numbers that will not agree with the first.

The place it matters: plain text, signatures, and anything a person might read out.

Why plain text breaks long URLs

RFC 5322 sets a line-length limit for internet messages, and mail clients wrap plain-text parts accordingly. A URL longer than the wrap width gets split across two lines, and a meaningful share of clients do not linkify the second half.

The reader sees a link that stops mid-path. Some of them paste both halves together; most give up.

A ten-character link cannot be wrapped. That is not a style preference, it is the difference between a working link and a broken one for whatever fraction of your list reads in plain text, which is higher than your HTML open rate suggests, because privacy-first clients render plain text by default.

A worked example

A weekly newsletter with 11,000 subscribers moved its footer links to short codes and left the body links alone.

Footer: gc.mw/unsub for preferences, gc.mw/arch for the archive, gc.mw/@ed for the editor's bio page. Body links stayed as full URLs, rewritten by the sending platform as usual.

Two things came out of it. The archive link, which had never been separately measurable because it sat inside the platform's aggregate footer number, turned out to collect more clicks than the third story in a typical issue. It moved up the page.

And when the newsletter changed platforms four months later, the footer survived intact. The body links did not, every one of them had been rewritten through the old platform's tracking domain, and every archived issue now points at a hostname nobody pays for. Roughly 200 issues of internal links, dead.

Deliverability, honestly

Spam filters weigh the reputation of every domain in a message. A brand-new shortener domain has none. A shared public shortener has whatever its worst user did last week.

The safe pattern is a subdomain of a domain you already send from, go.yourbrand.com when you send from yourbrand.com, because it inherits some standing rather than starting at zero. Three custom domains are included on Business and Team, and the setup is two DNS records.

The unsafe pattern is registering a brand-new short domain the week of a big send. It will look, to a filter, exactly like a phishing campaign looks.

Warm it up: put it in low-volume transactional mail first, then in the newsletter footer, then in the body. Weeks, not days.

The signature, which nobody instruments

Count the emails you send in a year. That is the impression count on your signature link, and almost nobody knows its click count because nobody has ever given it a code.

One short link fixes both problems. gc.mw/@you in the signature gives you a number, and it gives you one field to edit when the booking page, the portfolio or the company site moves, which it will, and which otherwise means asking every colleague to update their signature.

Tag it once, properly

Source, medium, campaign, content, built without typos. Then either shorten the result or store the UTM fields on the short link so the visible URL stays ten characters and the tags still arrive through the redirect.

One warning specific to email: if your platform also appends its own campaign parameters, you can end up with two sets fighting over the same fields. Check the final landing URL in a real test send, not in the preview pane.

The branded links page covers stored UTM fields, and custom domain links covers the domain that keeps your archive alive after you switch platforms.