Setting up an automation rule

An automation rule has three parts: a trigger (the moment it watches for), optional conditions (who it applies to), and one or more actions (what it does).

Read as a sentence: when an AI Video Interview is scored, and the score is below 4, mark the candidate as Poor fit.

Before you start

Automations is enabled per organization. If you do not see Automations in the sidebar, ask your Goodfit contact to turn it on for your account.

Owners and admins can create, edit and delete rules. Recruiters can change whether a rule is running and can use the simulator, but cannot create or delete rules.

Build the rule

Go to Automations → New rule.

• Describe it in a sentence. In the Describe your automation box, type what you want, for example Reject anyone whose AI interview score is below 4 and send them the rejection email. Goodfit fills in the trigger, conditions and actions below.
• Answer anything it asks. If your sentence leaves something open, such as the exact score or which stage, it asks a short question with the answers ready to click. You can skip any question and set the field by hand.
• Check every field. Everything it fills in is editable. Nothing is saved until you publish.

You can also skip the sentence entirely and build the rule by hand using the Trigger, Conditions and Actions cards.

Test it before it does anything

When you publish, you choose how the rule starts: Dry run or Live now. Dry run is selected for you, and it is the safe way to start.

In dry run the rule is evaluated for real, against real candidates, and records exactly what it would have done. It does not move anyone, does not change any board status, and does not send any email. Each simulated step is recorded as Rule is in dry-run mode, actions simulated, not executed.

Leave a rule in dry run for as long as you need. Open Automations → Activity to see each time it was evaluated, whether the conditions matched, and which actions it would have taken.

There is also a simulator on the rule builder when you are working inside a single job. It answers a narrower question: how many of that job's applicants from the last 90 days already match your conditions. It looks at conditions only, not the trigger, so treat it as a rough sense of scale rather than a prediction of how often the rule will fire.

Take it live

When the dry run looks right, open the rule's menu in the list and choose Go live. From that point the actions really happen.

You can also pick Live now when you publish. The rule then acts from the first match, so you are asked to confirm that it acts on real candidates. Editing a live rule keeps it live unless you switch it to dry run when you save.

You can move a rule back to dry run at any time, or disable it entirely. From the menu, a disabled rule comes back as a dry run.

Who it applies to, and from when

Rules watch forward only. A rule starts watching from the moment you publish it. Candidates already sitting in your pipeline are not re-evaluated and nothing is applied retrospectively.

There is one exception worth understanding. The Time in stage trigger does count candidates who are already in the stage, but the clock starts the day you publish the rule, not the day they entered the stage. A rule set to 30 days will not act on someone who has already been waiting 45 days until 30 further days have passed.

If a rule stops working

A rule is marked broken automatically if it points at something that no longer exists, for example a pipeline stage that was deleted. Broken rules stop running rather than behaving unpredictably. Open the rule, point it at a stage that exists, and put it back through dry run.

Limits

These exist to stop a misconfigured rule from running away:

• A single rule acts at most 500 times per day.
• Your whole organization acts at most 5,000 times per day.
• If one rule's action triggers another rule, that chain stops after 3 steps.

Dry run activity does not count toward these limits.