Link-shortener domain — the three categories that decide CTR

Most posts on link-shortener domains pretend there are two categories — generic shortener (bad), branded domain (good). The real picture has three, and the middle one is where most operators lose the most clicks. A short link on bit.ly/3xK9pQ is at least a domain people recognise. A short link on rebrand.ly/abc or shor.tl/spring or some YC-startup-of-the-month shortener is a domain people don't recognise — and unrecognised reads worse in 2026 than known-but-generic. The unfamiliar middle is the trap: senders pick a "modern" shortener thinking they've upgraded from bit.ly, and ship the click-through rate further down instead of up.

The data backs this up across roughly a decade of email A/B tests, social-share trackers, and SMS conversion logs. A custom branded domain — your own, on your own DNS — sits at the top. bit.ly sits in the middle. Unfamiliar third-party shorteners sit at the bottom. This post walks through what each category signals, what the click-through-rate data actually shows, and how to score whichever short link you're about to send before you send it. The widget below the data section lets you paste any short URL and see which category it falls into. The companion visual-essay — custom domain short links, the case in five charts — pulls the CTR, trust-survey, TLD-phishing, SMS-deliverability, and compounding-value data into one chart-driven walkthrough.

Why three categories, not two#

The naive frame is "branded vs unbranded". That misses the recognition axis entirely. Two unbranded shorteners can perform very differently if one of them is a domain the recipient has seen 500 times before and the other is one they've seen zero times.

bit.ly is the canonical known shortener. It's been around since 2008, it's processed billions of redirects, every spam filter has a model for it, every recipient has clicked through at least one bit.ly link in their life. The recipient may not love it, but they don't fear it. It reads as "shortener I've seen before". That's a working trust signal even though it's a weaker one than a brand they actually recognise.

The unfamiliar third-party shortener is the inverse. The domain looks like a typo. The recipient has never seen it. Spam filters have weaker models for it because the volume isn't there. There's no track record either way. The default human reading is "this is suspicious".

A custom branded domain — yourbrand.link/spring, news.yourbrand.com/may, whatever — flips the script. The domain itself carries the brand's existing trust. Recognition is automatic if the recipient knows the brand. If they don't, they at least see a real-looking commercial domain that follows the same shape as everything else in their inbox. If you want to feel the difference, draft a slug against a branded domain in the short-link generator and compare its URL preview to the same destination on a generic shortener.

The directional pattern is consistent across every channel where the URL is visible to the recipient before they click — email, SMS, social posts with bare URLs, iMessage rich previews, Slack unfurls. Channels where the link is hidden behind anchor text (HTML emails, blog posts) muffle the effect because the recipient never sees the domain at all. The same three categories drive the scan-to-click decision when the link is embedded in a QR — covered in why the QR code domain matters as much as the design, since the camera-app preview banner functions as another "URL visible before click" surface.

What the click-through-rate data actually shows#

The cleanest data point on branded vs generic comes from the canonical Bitly Brand Health study, which reports an average click-through-rate lift of about 34% when senders move from generic shorteners to a branded custom domain. That number gets cited everywhere; it's been replicated by smaller studies inside ConvertKit, Buffer, and the BrightEdge B2B-traffic dataset. The 25 to 40 percent range in the figure above brackets it.

The unfamiliar-middle data is messier because it depends on which unfamiliar shortener you're looking at. Internal A/B tests inside email-platform vendors have measured -10 to -25 percent CTR for short-link domains the recipient hasn't seen before, compared with bit.ly. Two factors drive the gap. First, recipients hesitate at the unfamiliar URL — the hover-preview surface in iMessage, Slack, and modern desktop email clients shows the resolved domain before the click, and an unrecognised domain triggers more abandonment than a known one. Second, spam filters classify low-volume third-party shorteners more aggressively than bit.ly, which has built up a reputation profile that lets some legitimate mail through that fnxr.co or shor.tl wouldn't.

The "suspicious TLD" category at the right tail is a real one. A short link on .xyz, .click, .tk, or a single-letter ccTLD reads to recipients as scammy regardless of whether the underlying shortener is reputable. It also catches a heavier spam-filter penalty than a .com / .co / .link shortener does. If you've shopped a domain for short-link use and the only available short option is on one of those TLDs, the math is usually that paying $25 a year for a .co works out cheaper than the CTR cliff on a "free" .xyz.

Why the unfamiliar middle is the trap#

Operators end up in the unfamiliar middle for two reasons.

The first is the migration path. They started on bit.ly because everyone does. They got tired of bit.ly's pricing or restrictions and looked for alternatives. They picked a smaller shortener — Rebrandly, Short.io, Dub, T2M, whatever — without setting up a custom domain on top of it. The smaller shortener's default redirect host is something like rebrand.ly/abc or shor.tl/spring. The recipient has never seen that domain. CTR drops, and the operator usually doesn't know it dropped because they're comparing absolute numbers across products instead of running a controlled A/B.

The second is the "free domain" trap. Many shorteners offer a free shared domain — dub.sh, lstu.fr, is.gd — as a way to get started without a custom-domain setup. The free domain looks slick. It also reads to recipients as "the latest sketchy URL pattern I've never seen before". The CTR on a free shared domain is structurally lower than on either bit.ly or your own domain. Shorteners don't volunteer this number because it's the upsell hook for their custom-domain tier.

The honest conclusion: if you're going to use a smaller shortener at all, you have to bring your own domain to it. Otherwise you've moved off bit.ly into a category that performs worse, not better.

Score any short link in fifteen seconds#

Pick the shortener you're about to ship a campaign on, run the inputs through the widget below, and get a directional read on which of the three categories it falls into. The widget doesn't measure CTR — that needs a real A/B test in your own audience. It measures the recognition and trust factors that drive CTR, and pattern-matches against the categories above.

A 4/5 or 5/5 puts you in the branded category. 2-3 means you're at the bit.ly baseline. 0-1 means you're in the unfamiliar middle and the campaign is leaving clicks on the table that you can recover by setting up a custom domain.

What "branded" actually requires#

Three things, in this order.

A domain you own. Not a subdomain of the shortener's domain. Not yourname.bit.ly or yourbrand.dub.sh. Those still resolve to the shortener's host, not yours, and the recipient sees the shortener's domain, not yours. The honest test is whether the URL bar after the redirect (or the hover preview before it) shows your domain.

DNS pointed at the shortener's IP or CNAME. Most shorteners give you a short setup string — a CNAME from your subdomain to their domain, or an A record to their IP. Propagation usually completes inside five minutes; the full walkthrough on this is in setting up a custom short-link domain. Cloudflare DNS is faster for propagation than registrar-default DNS but both work.

Auto-TLS issuance. The shortener should issue a Let's Encrypt certificate the first time someone hits HTTPS on your hostname. If the shortener requires you to upload your own cert, the setup overhead is high enough that operators skip it. Caddy on-demand TLS is the canonical pattern; most modern shorteners ship it; some older ones don't.

Operationally, the work is twenty minutes the first time and zero minutes thereafter. The CTR delta keeps paying every campaign for as long as you keep the domain.

Recognition is path-dependent#

One more wrinkle worth naming. Recognition is built over time, by repetition. The first three or four times your audience sees yourbrand.link they read it as "unfamiliar — what is this". By the tenth or twentieth send, they read it as "ah, it's them again". The CTR lift compounds over a quarter; it isn't fully present on the first send.

This means switching domains has a temporary CTR cost that the new domain has to grow out of. The cost is small — maybe 5-10 percent on the first send compared to where you'll be in three months — but it's real. The implication is "pick the right short-link domain once and stick with it", not "iterate on shortener domains the way you iterate on landing-page copy". You're not optimising the URL the way you optimise a CTA button; you're building recognition the way you build email-sender reputation.

If you're considering switching shorteners purely to chase a feature, factor in the CTR-rebuild cost. If the feature is critical (custom domain support that you didn't have before, real analytics, dynamic redirect — covered in bitly alternatives in 2026), the switch pays back. If the feature is cosmetic, the rebuild cost dominates.

The other side — when shortener domain doesn't matter much#

Two cases where the three-category pattern flattens out.

Anchor-text-only contexts. HTML emails where the link is hidden behind a "Read more" button. Tweets where the link is auto-rendered as a card and the user clicks the card image, not the URL. iOS Mail's read pane on long emails where URLs disappear into "see more" expansions. The domain isn't visible to the recipient before the click, so it can't influence the click. The CTR is driven entirely by the anchor text and the channel context.

Already-trusted channels. A signed-in customer in your own app, where every link goes back to your domain anyway. A private Slack channel where everyone trusts what gets posted. A community newsletter where the recipient has paid for the content. The trust scaffold is provided by the channel, not by the URL.

Outside those cases — cold-list email, public social posts, SMS, public-link sharing, support tickets, embedded marketing CTAs — the domain matters and the three-category pattern holds.

Related reading#

• Branded short links — why the domain matters — the broader two-category case for branded over generic.
• Branded short links, trust, and the click you lose — the trust-mechanic deep-dive with a CTR-loss calculator.
• Setting up a custom short-link domain — DNS, TLS, and verification step-by-step.
• Bitly alternatives in 2026 — the platform comparison that puts custom-domain support at the top of the criteria list.
• Short-link SEO — myths and facts about redirects — the SEO side of the same domain question.
• Custom domain — platform docs — the per-tenant wiring on the Linked.Codes side.