How to use Zapier Paths and Filters for conditional routing

Filter by Zapier: one condition. If the condition is met, the rest of the Zap runs. If not, the run ends. Use when: you want to ignore some trigger events entirely.

Paths by Zapier: branch the Zap into 2-5 paths based on conditions. Each path has its own action sequence. Use when: different conditions lead to different downstream actions.

Rule of thumb: if all branches share most of the same actions, use Filter inside individual actions or use Paths with overlapping logic. If branches diverge entirely, use Paths.

Cost: Filter steps consume 0 Tasks. Paths steps consume 0 Tasks. Only the action steps inside the matched branch consume Tasks. This makes Filter+Paths the cheapest way to control Task usage.

In your Zap editor, click the "+" between the trigger and the first action.

Search the action picker for "Filter." Choose "Filter by Zapier."

Choose "Only continue if..." (the default and almost always what you want).

Configure the condition: pick a trigger field, pick an operator (exists, equals, contains, greater than, etc.), pick or type a value.

For multiple conditions joined by AND: add them in the same Filter step. For OR logic: add a second condition group ("Or") — Zapier shows this as a separate row group.

Click "Test step." Zapier shows whether the current sample passes or fails the filter. If your sample fails but you expected it to pass, the condition logic is wrong.

Click "+" → search "Paths" → choose "Paths by Zapier."

You now see two starter branches: Path A and Path B. Each branch is a mini-Zap.

For each branch, click in and configure the condition first. Use the (field, operator, value) pattern.

Critical: make conditions mutually exclusive. If a record could match both Path A and Path B, only Path A will run (paths evaluate top-down). Hidden records will be missed by all paths.

Inside each branch, add the actions for that path. Each branch can have its own multi-step chain.

Add additional paths with the "+ Add a Path" button. Max 5 paths per Paths step on Pro; higher tiers allow more.

For a "catch-all / else" path: configure conditions so the last path matches anything not caught by earlier paths.

You cannot trust Paths logic without testing every branch with real data.

Generate one trigger event for each branch. For three branches (enterprise, paid, free), submit three real records, one per tier.

Open Zap → History. For each run, expand the Paths step and verify the chosen branch matches your intent.

For each chosen branch, expand the action steps and verify they ran. Verify destination apps received the correct records.

If a record went down the wrong branch, your condition logic is inverted or has a missing case. Fix in the editor and retest.

Default Paths behavior: if a record matches no branch, the Zap ends without running any actions. No error, no log entry indicating "skipped."

This is the silent-failure mode most DIY setups never catch.

Add a final Path with a permissive condition — usually "trigger field exists" or "Path field does not equal A and does not equal B."

Inside this catch-all path, add a single Slack or email notification: "Zap [name] received a record that matched no path. Investigate."

Now any gap in your condition logic becomes visible within an hour instead of invisible forever.

Before turning the Zap on at full volume, check current Task usage in Zapier → Settings → Usage.

Estimate expected Task usage after Paths/Filters: count actions per branch, multiply by expected branch volume.

After 7 days live, compare actual vs. estimated. If the reduction is smaller than expected, your filter conditions are too permissive.

Common pattern: a 5-step Zap firing 1,000 times/mo at 5,000 Tasks/mo drops to 2,000 Tasks/mo after adding a Filter that skips 60% of events that should not trigger downstream actions.