Automation

Zapier Tasks Explained: How Many You Actually Need

What counts as a task in Zapier, what's free, and how to estimate your real monthly usage before you pick a plan — with the arithmetic done for three workflows.

Zapier’s free plan gives you 100 tasks a month. The next plan up gives you 750. Neither number means anything until you know what a task is — and most people guess wrong in the same direction, assuming one automation run equals one task.

It doesn’t. But the good news is that Zapier’s counting is unusually forgiving: several things you’d assume are billed aren’t billed at all, including the ones that fail.

Here’s the rule, then the arithmetic for three real workflows.

What Counts as a Task

ElementCost
TriggerFree, always
Successful action1 task
Filter or Path stepFree
Action that errors or haltsFree
Steps skipped because a filter stopped the ZapFree
Search step, “don’t proceed if nothing found”Free when nothing is found
Search step, “proceed if nothing found”1 task

Two things follow from this, and both are more generous than people expect.

Filters are free, so filter early. A filter costs nothing and prevents every step after it from running — and those skipped steps cost nothing either. Putting a filter immediately after the trigger is the single most effective way to cut task usage. There is no reason to hesitate about adding one.

Failures are free. If an action errors, it doesn’t consume a task. A Zap that breaks and fails fifty times overnight has cost you nothing but the broken automation. Worth knowing before you panic about a runaway Zap.

Zap Runs vs Tasks

These are different numbers and the distinction matters when you read your usage dashboard.

A Zap run is one trigger event — one form submission, one new row, one payment. A task is one successful action inside that run.

So a Zap with two actions, triggered ten times, is 10 runs and 20 tasks. Your plan limit is measured in tasks.

Three Workflows, Priced

A simple intake — 4 tasks per run

Form submission → create CRM record → send acknowledgement → create a task → notify your phone. The trigger is free; four actions cost four tasks. At 20 enquiries a month: 80 tasks. That fits the free plan’s 100 — except the free plan won’t let you build it, which we’ll come back to.

A filtered intake — 1 task per run, most of the time

Same workflow, but with a filter after the trigger that only continues for enquiries over a certain budget. If three of twenty submissions pass the filter, you pay for those three runs only: 12 tasks a month instead of 80. The seventeen filtered-out runs cost nothing.

That’s an 85% reduction from one free step.

A daily sync — the expensive shape

Every morning, fetch yesterday’s orders and update a spreadsheet row for each. Fifteen orders means fifteen successful actions: 15 tasks per run, about 450 a month. Zapier bills per successful action regardless of how the data arrived, so list-processing is where its costs climb too — just less steeply than Make’s operations, where every downstream module runs per item.

The Free Plan Problem

This is the part that decides whether Zapier is usable for you at all.

The free plan’s headline limit is 100 tasks a month, which sounds workable. The real limit is that Zaps are capped at two steps — one trigger, one action.

“When a form is submitted, add a row to a spreadsheet” fits. “When a form is submitted, add a row, send an email, and notify me” does not. Almost every automation in our roundup of automation tools for solopreneurs needs more than one action.

So the free plan is a way to confirm a connection works, not a way to run a business. Budget for the paid tier from the start, or start on a tool whose free plan allows multi-step workflows — which is the practical argument in our Zapier vs Make comparison.

How to Estimate Your Own Usage

  1. Count the actions in each Zap. Ignore the trigger and any filters — they’re free.
  2. Estimate how often it runs per month, and subtract the runs a filter will stop.
  3. Multiply, then add about 20% for testing and the runs you forget about.

If a step processes a list, count one task per item rather than one per run. That’s where estimates usually break.

Four Ways to Cut Task Usage

  • Filter as early as possible. It costs nothing and everything downstream of a stopped Zap is free. This is the biggest lever by a wide margin.
  • Use Paths instead of separate Zaps. Path steps are free, and one Zap with branches usually costs fewer tasks than three Zaps each doing their own lookups.
  • Set searches to “don’t proceed if nothing found.” That configuration costs nothing when the search comes up empty; the other one costs a task every time.
  • Collapse redundant actions. Two steps writing to the same record where one update would do is one wasted task on every single run, forever.

Frequently Asked Questions

Solotimize is written by a one-person business, for one-person businesses. Recommendations are judged on what fits a solo operator’s time and budget, never on commission — and we always tell you whether we’ve used a tool or researched it. Pricing referenced above was accurate at the time of publication; always confirm current plans with the vendor.


No. Triggers never consume tasks, no matter how often they fire or how much data they return. Only successful actions count. This is a meaningful difference from Make, where the trigger costs one operation per run — on a simple linear workflow, that single difference is most of the gap between the two tools’ counts.


No. Actions that error or halt don’t count toward your usage, and steps that never ran because a filter stopped the Zap don’t count either. A Zap that breaks and fails repeatedly costs you nothing in tasks. It’s still worth fixing quickly, but the damage is to your workflow, not your bill.


Most one-person businesses running event-driven automations — form submissions, payments, new clients — land somewhere between 200 and 800 tasks a month, which the entry paid plan covers. You’ll exceed that if you run scheduled syncs that process lists, because those bill one task per item rather than per run. Estimate with the method above rather than guessing, and remember that a filter placed early can cut a workflow’s usage by most of its total.


Keep reading