Custom domain short links — the case in five charts

Most arguments for custom domain short links hand-wave the data. "Branded gets more clicks" is true, but the receipts live in a dozen separate places — agency case studies, deliverability papers, browser-vendor security reports, ICANN abuse statistics, search-engine link-graph research. Pulled into one place, the case sharpens. There are five distinct mechanisms working in parallel, each measurable, each pointing the same direction. Five charts, five sections, one conclusion.

This post is the visual-essay companion to the longer argument piece — branded short links, why your domain beats bit.ly — and the data deep-dive that ranks the three shortener categories in link-shortener domain CTR data. Read those if you want prose. Read this if you want the case in five charts you can paste into a deck and walk a stakeholder through in eight minutes.

Chart 1 — Click-through rate by domain category#

The headline number anyone arguing for custom domains has to lead with. CTR isn't the only reason to switch, but it's the one that turns up on a P&L.

The shape of the chart is what matters. Branded sits at the top. bit.ly is the reference. The unfamiliar middle — every "modern" shortener that doesn't have a custom-domain layer over it — sits below the reference, not above. Operators who migrate from bit.ly to a smaller shortener thinking they've upgraded usually ship their CTR down. The full breakdown on why that's structurally true is in link-shortener domain — the three categories that decide CTR. The mechanism is recognition: bit.ly has been clicked billions of times, every spam filter has a model for it, every recipient has seen it. A fresh shortener has none of those signals. A custom domain on your own brand has the signal a fresh shortener lacks, plus the recognition bit.ly built — yours instead of theirs.

The 34% number is the most-cited figure. The honest range is 25 to 40 percent depending on channel and audience. For cold-list email and SMS the lift sits at the top of the range. For already-trusted channels (signed-in users, paid newsletters) the lift collapses to single digits because the trust scaffold is provided by the channel.

Chart 2 — Preview-bar trust by URL shape#

The second mechanism is what recipients see before they click. Every modern messaging surface — iMessage, Slack, Gmail desktop, Outlook desktop, Apple Mail on iOS — shows a resolved-URL preview either inline or on hover. The preview is the moment recognition fires or doesn't.

The seven-row gap is the same trust mechanic the branded short links, trust, and the click you lose deep-dive walks through in prose. Three things stack: domain familiarity (have I seen this before), TLD legitimacy (is this a real-looking commercial suffix), and slug shape (does it read as a campaign word or a random hash). A custom domain wins all three. Generic shorteners win one. Unfamiliar third-party shorteners win zero. The score is what shows up in CTR.

What the chart doesn't capture is the timing. The preview bar fires before the click — recipients spend somewhere between 200 and 800 milliseconds reading the resolved domain on hover. That window is short enough that conscious analysis doesn't happen; recognition either fires or it doesn't. Which means you can't talk people into trusting your unfamiliar shortener with a longer email body. The domain is doing the work the body never gets to do.

Chart 3 — Phishing classification rates per TLD#

The third mechanism is the spam-filter and browser-warning side of the same trust problem. ICANN and the Anti-Phishing Working Group publish quarterly abuse statistics by TLD. Some TLDs concentrate phishing at rates 20 to 50 times higher than others. Spam filters and browsers read those statistics directly into per-TLD reputation scores.

The practical lesson is in the QR code security and quishing post on the scan side, but the spam-filter version of the same lesson lives on the email side. Pick the wrong TLD for your short-link domain and you ship every campaign starting at the bottom of the deliverability curve. The gap between a .com short-link domain and a .xyz short-link domain is roughly 8 percentage points of phishing base rate — which the filter reads as a 20-times-worse starting score. Some of that gets recovered as your domain builds sender reputation, but you're climbing out of a hole the .com domain never falls into.

The shopping list is short. .com, .co, .net, .link, or a country-code TLD (.de, .fr, .uk) that matches your audience. Avoid .xyz, .click, .top, .tk, single-letter ccTLDs, and anything ICANN has flagged as a "high-abuse" gTLD in its monthly registry abuse reports. The annual price difference between a "good" TLD and a "bad" one is usually under $30. The cost of the deliverability hit is much larger than $30.

A custom short-link domain is the only piece of campaign infrastructure that gets re-read on every send for years. Get it right once and it compounds. Get it wrong once and it bleeds.

Chart 4 — SMS deliverability before and after Apple's spam filter#

The fourth mechanism is platform-specific but big. In late 2024 Apple shipped iOS 18's "Suspected Spam" filter, which auto-routes SMS from unknown senders containing certain URL patterns into a separate Junk folder. The filter learned aggressively over the following twelve months. Generic-shortener SMS got hit hard. Branded-domain SMS mostly didn't.

The mechanism is simple. Apple's filter is conservative on URLs from domains it can't fingerprint. A branded short-link domain looks like a normal commercial host the recipient might already know. A generic shortener is a pattern Apple has trained an aggressive classifier on. An unfamiliar shortener looks like everything else the filter is trying to catch.

For SMS-heavy campaigns this is existential. The 30-point delivery gap between branded and unknown shorteners means a 10,000-recipient campaign sees roughly 5,500 messages reach the inbox tab instead of 2,800. Click-through-on-delivered stays at ~10%. So delivery is the bottleneck and the bottleneck has just doubled the gap between domain categories. The deeper version of how SMS message-segment cost stacks with the domain-trust effect is in URL shortener for SMS — keep the message in one segment.

Chart 5 — Compounding domain authority over years#

The fifth chart is the one most "branded vs generic" arguments skip. A short-link domain doesn't just perform on day one — it compounds. Every send, every share, every backlink to your branded short URLs adds to the domain's reputation in the link graph. Generic shorteners can't compound that value to you because the reputation accrues to the shortener's host, not yours.

The lines diverge because every signal the domain accumulates only matters if it accumulates to something you own. Email-sender reputation, browser anti-phishing whitelists, manual user recognition, backlinks from third parties citing your short URLs, search engines occasionally indexing the slug instead of the redirect target — all five effects compound on a domain you control. None of them compound to you on a generic shortener. Three years in, the operator who chose a custom domain in year one is sending into a fundamentally different deliverability landscape than the one who didn't. The infrastructure-ownership version of the same argument is in owning your link infrastructure — same compounding logic applied to the full redirect stack instead of just the domain layer.

The number that matters here isn't the year-5 multiplier. It's the gap. By year three the branded domain is sending into a context that's 2 to 3 times more receptive than the generic shortener. That gap doesn't show up in a single-campaign A/B test. It shows up in the cumulative click count of the operator who switched early.

A score above 80 puts the link in the top category — recipients pattern-match it as legitimate and the CTR runs at the high end of the published range. A score in the 30 to 55 band is recoverable but bleeding. A score under 30 is a domain that shouldn't be shipping campaigns. The score is directional, not a guarantee — your actual CTR is a function of the audience and the message too — but the four inputs above are the ones that drive most of the variance the data above captures.

Where the five charts converge#

Pull the five mechanisms together and the case writes itself. Click-through is higher on a branded domain. Trust survey results are higher on a branded domain. TLD-based phishing classification is lower on a branded .com than on a generic .click. SMS deliverability survives Apple's filter on a branded domain and doesn't on a generic shortener. And the value of a branded domain compounds year over year while a generic shortener accrues that value to someone else.

The implication isn't "branded short links are nice to have". It's that the operator running campaigns on bit.ly or rebrand.ly in 2026 is structurally behind the operator who set up a custom domain in 2023. The gap widens every send. The setup work — about 30 minutes to point a DNS record at a short-link platform and let it issue a Let's Encrypt certificate — is small enough that the only honest reason to skip it is "I didn't know it mattered this much". The custom domain platform docs walk through the per-tenant wiring. The plain-English version of what a "vanity URL" even is, since the term gets used interchangeably with branded short links, is in what is a vanity URL.

If you want to test what one of these branded URLs looks like on your own domain before committing, the short-link generator creates one for free against the Linked.Codes default host. Move it to your own domain later when you're ready to keep the recognition value compounding to you instead of to us — the lifetime tier on the pricing page includes custom-domain support and auto-TLS without the per-month rent that makes the five-year compounding chart above land badly on a SaaS subscription budget. The SEO side of the same migration — and the myths about whether branded redirects help or hurt search — is in short-link SEO myths and facts. The full step-by-step on the move is in setting up a custom short-link domain.

The five charts aren't five independent arguments. They're five views of the same underlying mechanism: recognition. Recipients trust what they recognise; spam filters trust what they recognise; search engines and browser warnings reward what they recognise; and the only thing you can build recognition on is a domain you own. The custom-domain decision is the single piece of campaign infrastructure where the right answer is genuinely the same for almost everyone — the exception list is "single-use throwaway link" and "internal team Slack ping", which is to say, nothing that ships to customers.

Sourcesshow citations

• Bitly Brand Health study — referenced across industry coverage of branded vs generic short-link CTR — https://bitly.com/blog/
• Anti-Phishing Working Group quarterly phishing activity trends reports — https://apwg.org/trendsreports/
• ICANN Domain Abuse Activity Reporting (DAAR) — https://www.icann.org/octo-ssr/daar
• Spamhaus Top 10 Most Abused TLDs — https://www.spamhaus.org/statistics/tlds/
• Google Safe Browsing transparency report — https://transparencyreport.google.com/safe-browsing/overview
• Apple Developer documentation on iOS Messages filtering — https://developer.apple.com/documentation/sms_and_call_reporting
• Litmus State of Email engagement and deliverability benchmarks — https://www.litmus.com/resources/state-of-email
• HubSpot Marketing Statistics — CTR benchmarks by channel — https://www.hubspot.com/marketing-statistics