Notifications

When something happens on your account, Routy records it once and then delivers it to the places you asked for. An account integration stops collecting, a link stops redirecting, a postback is rejected, a custom domain's certificate check fails, a CSV import is ready for review. Each of those is a notification type, and each type has its own set of channels you can switch on or off.

There are four channels: the bell in the portal, a browser push notification, email, and a Slack channel your team already watches. They carry the same event and differ in how fast they reach you and how hard they are to ignore. The bell is the record, so whatever else you switch on or off, an event Routy records is visible in the portal. The other three push it to you sooner.

This is about telling people. If you want your systems told, so software can react without anyone reading anything, that's operational webhooks, and the same events feed both.

• Channel control per notification type. Settings > Notifications lists every type with the channels it can use. Switch a channel on or off per type, or reset a type to the Routy default. Your settings are yours, so changing them doesn't change a colleague's.
• Live alerts while you have Routy open. They appear the moment the event happens, without a refresh, and carry a level: error, warning or info. Alerts follow the account you have selected in the switcher, plus anything addressed to you personally.
• Email by default for the things you may have to act on, and not for the things Routy already handled and only wants on the record.
• One Slack webhook for the whole account. Paste an incoming-webhook URL once in your notification settings and every type with Slack switched on posts to that channel. You don't set up a webhook notification type by notification type. Subscription and billing alerts can reach Slack as well.
• Slack cards that name the account they're about, rather than identifying it by its internal number alone, with a colour bar by severity.
• A recurring fault as one entry carrying a count and the time it last happened, instead of twenty copies of itself. An entry that fired once looks exactly as it did before.
• A notification settings link in the footer of every notification email and in every Slack message, so you can change what sent it, or switch it off, from the message itself.
• Browser push on desktop and mobile, so an alert reaches you with Routy closed. Each device has its own subscription, and an alert goes to every device you enabled.

Two things your own preferences deliberately don't control. Slack is one, because a Slack channel belongs to the team and one person's setting shouldn't silence a post everyone else relies on, so Slack is configured for the account. Webhooks are the other, because those are your systems' subscriptions rather than your inbox, and a preference you set for your own reading shouldn't disable an integration.

Choosing what you get

There are two levels. Your own preferences answer what you want to hear about. Account settings answer what this account does by default, and an account admin sets them. Routy resolves a notification type in this order and takes the first match: your own preference for that type, then the account's preference, then the type's own Routy default.

A type nobody has touched follows the Routy default. An account admin can change the default for everyone who hasn't set a preference of their own, and a preference you did set survives that change.

Some notifications can't be switched off. Account confirmation, password reset and the like are part of using the account rather than something you opted into, and the preference screen locks them.

Setting up Slack

Paste an incoming-webhook URL from Slack into the account's Slack setting and enable it. Test it before you save: Routy posts a real verification message to the URL you pasted, so you find out then whether it works instead of discovering months later that nothing was ever delivered. Once saved the URL is shown masked, so to change it, paste the whole new URL rather than editing the masked one.

If you want one particular type going somewhere else, a payments channel for instance, put a different webhook URL on that type's own preference. That override wins for that type only, and everything else keeps using the account webhook.

A type with Slack switched on and no webhook configured anywhere posts nothing. That's the most common reason for "Slack is on but we get nothing".

Which emails arrive by default

The default is failures, not everything. Email is the channel people can't filter without effort, and a mailbox full of "we handled this for you" notices is a mailbox people stop reading, including on the day something is broken. So Routy emails you by default when there may be something you have to act on:

• an account integration stopped collecting, through wrong credentials, a blocked IP, or an action the affiliate program is waiting on
• a link stopped redirecting, or clicks on it are being rejected
• a custom domain's DNS, certificate or app check failed
• a quota is exhausted or is blocking something
• a postback was rejected
• payment, subscription and invoice events

Six informational types changed on 22 August 2026 and are now bell-only unless you ask for them: Link Changed, Account Integration: Slower Tier, Traffic Sources Auto-Archived, Brand Links Auto-Archived, Duplicate Channel Data Resolved, and Account Integration: Temporarily Paused. All six still raise an alert in the portal, and all six will mail you correctly the moment you switch Email on for them.

Email needs a confirmed address on your user and an active account. It goes to the account's enabled, email-confirmed users who haven't opted out of that type, except for billing and ownership notices, which go to the designated contacts.

Browser push

Push works in Chrome, Firefox, Edge and Safari 16 or later on Windows and macOS, and in Chrome, Firefox and Edge on Android. On iOS it works in Safari 16.4 or later, and only after you add Routy to your home screen and open it from there. Chrome, Edge and Firefox on an iPhone can't deliver push at all: Apple requires every iOS browser to use Safari's engine, and that engine doesn't give a third-party browser the Push API.

Enable push under Settings > Notifications, allow the browser prompt, then switch the Push column on for the types you want. A Test Notification button in the same screen checks the setup. Delivery uses VAPID, with no third-party service such as Firebase in the path, and the subscription stores an endpoint URL for your browser, encryption keys, and optionally your browser and device type for troubleshooting.

Push needs the device on and connected to the internet. Something raised while the machine is off arrives once it's back.

When the same problem keeps happening

Two separate rules keep a repeated fault from filling your list. Occurrences that share an event identifier are deduplicated, so while one is open another occurrence updates it rather than adding a row, and the entry shows how many times it has happened and when it last did. Separately, a notification type can carry a minimum gap in seconds, and an occurrence inside that window is skipped rather than pushed at you again.

Both rules drop the repeat rather than sending you a digest of it. If you're counting occurrences, the count on the entry is the number to read, not the number of messages you received.

Most of what Routy notices is something you'd otherwise find as a report that went quiet. An account integration that stopped signing in keeps producing reports that look fine and have no new rows in them, and a rejected postback looks exactly like a conversion that never happened. Both are email types by default, because somebody has to go and fix something.

The two-level setup exists for the usual split in a small team. The person handling tracking and the person handling payouts want different types in their inbox, and both want the broken-integration alert in the Slack channel where the team already talks. Account-level Slack with per-person email gets that without either of them triaging the other's alerts.

Email defaulting to failures only is there to head off the "ignore all Routy notifications" reflex. A channel you can't filter gets muted wholesale once it starts carrying routine notices, and then the alert that mattered is muted along with it. Keeping the tidy-up notices in the bell, where scrolling past them costs nothing, is what keeps the inbox worth opening. Switch them on if you want them and they'll arrive.