Solutions

The jobs people hire it for.

Every one of these starts the same way: a link that's already out there, and a destination that needs to change. Here's what that looks like, and what it saves.

Start free How the platform works

canary & cutover

Roll out gradually, roll back instantly

Outcome: Smaller blast radius. A bad release reaches 10% of traffic instead of all of it, and undoing it takes one API call.
The problem: Moving traffic to a new endpoint is all or nothing, and the destination may not be yours to put a load balancer in front of.
How it works: Point a route at both endpoints, weight them 90/10, and ramp as confidence grows. Change the weights by API or in the dashboard.

targeting

Send each visitor to the right place

Outcome: One link, fewer wrong landings. Print it, post it, or embed it once.
The problem: You publish a link to people on iPhones, Android phones, and desktops, in several countries and languages, and each needs a different destination.
How it works: Select the target by country, language, device, operating system, browser, header, or query parameter. Visitors who match nothing go to a default you choose. Available on paid plans.

distribution

Distribute releases across mirrors

Outcome: No dead download links. You can repoint a published link without shipping a new build.
The problem: Once a download URL is in a README, a package manifest, or an email, you can't change it.
How it works: Publish one Divertly link. Spread it across mirrors or CDNs with weighted or random selection, and change where it points whenever you need to.

programmatic

Generate links from your product

Outcome: Per-customer links that don't hit a ceiling. You pay for clicks, not for how many links exist.
The problem: Marketing link tools cap link counts and price per link. Building your own means owning SSL, abuse handling, and uptime.
How it works: Create routes from your application through the API: one per customer, per tenant, per record, on your own domain. Routes and targets are unlimited on every plan.

experiments

Split traffic and measure what happens

Outcome: Evidence before commitment. You see what each destination does before you pick a winner.
The problem: You have two candidate destinations and no clean way to send each a share of the same link.
How it works: Weight the targets, send the link out, and read the click metrics per route.

migration

Migrate between providers without a cliff

Outcome: A reversible move. If the new provider misbehaves, you shift the weight back.
The problem: Cutting every link over to a new provider at once puts the whole migration on one bet.
How it works: Move traffic from the old destination to the new one gradually by shifting target weights. Nothing about the destinations has to change.

branded links

Run branded links on your own domain

Outcome: Links that look like yours. Customers see your brand, not ours.
The problem: Registrar fiddling, certificates, and DNS work for every campaign or customer.
How it works: Point a CNAME from your link domain and verify it. Full nameserver delegation is coming soon.

Pick one and try it free.

Weighted and random routing are on the Free plan, so most of this works on day one.

Start free Read the docs