Short links that carry your brand, not ours

A link is created in one field and refined in one dialog

Paste a destination on the overview screen and you get a short link on your default domain immediately, with a copy button next to it. That is the path for the ninety per cent of links that need nothing special. When a link does need more, the same record opens in a dialog with every option grouped into tabs, so a campaign link and a throwaway link are created by the same tool.

Slugs are yours to control. Type one and availability is checked against the chosen domain while you type; leave it empty and a short random slug is generated. Slugs are unique per domain and compared case-insensitively, so a visitor who types your link in capitals still lands on the right page. Each link also carries a title, a private note, tags and an optional folder — the metadata that keeps a list of ten thousand links searchable a year later.

The redirect status code is a per-link setting: 301 for permanent moves that should be cached by browsers and search engines, 302 for campaigns you intend to repoint, 307 and 308 when the request method must survive the hop.

Every behaviour lives on the same link record

Options are not scattered across separate products. One link can expire on a date, ask for a password, open a native app on mobile, route European traffic to a localised page, split the rest across two variants, fire a retargeting pixel and still append campaign tags to whichever destination wins.

The order in which those rules are evaluated is fixed and documented, because guessing is what makes link tools frustrating: expiry, then password, then social preview for crawlers, then deep links, then targeting rules, then the A/B split, then pixels, then the plain destination. UTM parameters are appended to the URL that actually wins, not only to the base one, so a targeted or split visit never loses the campaign tagging it was created for.

• ✓Expiration — a date and either a fallback URL or a branded expired page
• ✓Password — bcrypt-hashed, unlocked by a signed cookie that lasts fifteen minutes
• ✓Targeting — ordered rules on country, device, operating system and language
• ✓A/B — weighted variants that always sum to 100 per cent
• ✓Deep links — separate iOS and Android destinations
• ✓UTM — parameters appended to the winning destination
• ✓Open Graph — your own title, description and image for social previews
• ✓Pixels — retargeting tags fired on an interstitial before the redirect
• ✓QR — a scannable code for the same link, in PNG or SVG

Bulk work that does not turn into spreadsheet archaeology

Campaign links arrive in batches, so the list handles them in batches. Select rows and archive, restore, tag, untag, delete or export them in one action. Export writes a CSV of the current selection or of the whole filtered list, with tags and folders included, so a report for a client is two clicks rather than a query.

Import runs the other way: upload a CSV with destination, optional slug, title, tags and domain, review a per-row validation preview where broken rows are flagged inline, and create up to a thousand links per batch. The result comes back as a CSV with the short URLs filled in, ready to paste into whatever scheduled the campaign. Exports from Bitly, Short.io and Rebrandly are recognised by a mapping step, which is what makes moving an existing library realistic.

Filters are part of the address bar. Search text, domain, tag, folder and archive state all live in the URL, so a filtered view can be sent to a colleague as a link and it opens exactly as you left it.

The redirect path is the part that must never fail

Links are served by a worker at the edge from a key-value cache, not by a database query on the critical path. A change made in the dashboard reaches the edge in tens of milliseconds, and the cache is written before the request that missed it is answered, so the next visitor is served from cache rather than from a race.

Misses are handled defensively. An unknown slug is cached as negative for a short window so that a crawler hammering a dead URL costs one lookup instead of thousands, and that miss path is rate limited per address. If the dashboard is briefly unreachable, the edge answers 404 but refuses to write a negative entry — a temporary outage must not turn working links into dead ones.

What the visitor sees when something is wrong is still your brand: the not-found page, the expired page, the password prompt and the rate-limit page are rendered by the edge using the logo, colour and support address attached to the domain. Each domain also decides what its root and its unknown slugs do — return 404, or redirect to a page of your choosing.