Zapier is the tool many stores reach for first, because it connects thousands of apps without code. The surprise tends to come a few months later, when a store that started with a handful of Zaps finds it has exhausted its monthly allowance on the 18th. The cause is rarely a bug. It is arithmetic: orders times steps.
This guide explains how Zapier counts usage according to its published rules, shows how to estimate what your store will use, and describes the design changes that reduce it. It does not quote plan prices, because they change; you should check Zapier’s pricing page when you do your own sums. It pairs with Zapier vs Make vs n8n for Shopify, ecommerce automation cost and n8n with Shopify. Counting rules were read from Zapier’s help centre on 20 October 2026 and can change.
How Zapier counts tasks
According to Zapier’s help centre, a task is any successful action that runs in a Zap. The rules that matter most to a store are these:
| Step type | Uses a task? | Note |
|---|---|---|
| Trigger | No | Triggers are free, including polling checks |
| Action step that succeeds | Yes, 1 task | Each successful action counts |
| Filter | No | A filter that stops the Zap also prevents later steps from counting |
| Paths | No | Only the steps that run in the chosen path count |
| Formatter, Delay, Looping, Digest, Storage | No | Built-in utilities are excluded |
| Failed or halted step | No | Errors are not counted, and steps that never run are not counted |
| Search action | Depends | Counts one task when set to proceed if nothing is found; none when set to stop |
| Replayed steps | Yes | Steps rerun in a full replay count again |
| Some special products | Varies | Code, AI and other tools can use different amounts; check the current list |
Two consequences follow. First, a Zap that runs ten times and has four successful action steps uses forty tasks. Second, a filter at the start can make a difference: if only 30 percent of runs pass the filter, the later steps count for only 30 percent of runs.
Estimating usage for a store
- List every Zap and its trigger.
- Estimate how many times each triggers per month, using real order and event counts.
- Count the action steps that execute on a typical run, after filters and paths.
- Multiply runs by steps for each Zap.
- Add them up, then add a margin of 20 to 30 percent for growth, replays and mistakes.
- Compare with plan limits and overage rules from Zapier’s current pricing page.
| Zap | Runs per month | Action steps that run | Tasks |
|---|---|---|---|
| New order to spreadsheet | 1,500 | 1 | 1,500 |
| New order to warehouse (create, tag, notify) | 1,500 | 3 | 4,500 |
| High-value order alert (filter passes 8%) | 1,500 triggers, 120 pass | 1 | 120 |
| Tracking update to customer email | 1,500 | 2 | 3,000 |
| Low-stock alert (filter passes 5%) | 3,000 checks, 150 pass | 1 | 150 |
| New review to helpdesk | 200 | 2 | 400 |
| Refund to accounting | 120 | 2 | 240 |
| Total | 9,910 |
The figures are illustrative, but they show the shape: the Zaps that run on every order dominate, while filtered alerts hardly register. A store doing 1,500 orders a month with six Zaps uses nearly ten thousand tasks, and doubling orders roughly doubles the usage. Revenue-driven growth therefore translates directly into automation cost.
Where stores waste tasks
- Filters placed late. A filter after three actions saves nothing; put it first.
- Many tiny steps. Several “update field” actions that could be one update.
- Duplicate Zaps. Two Zaps doing the same job after someone copied one for testing.
- Looping over big lists with an action per item, when a bulk action or a database operation exists.
- Replays and failed-run recovery done by rerunning everything.
- Notifications for every event instead of a daily digest.
- Using action steps for formatting that a free Formatter step could do.
- Syncing data that rarely changes on a tight schedule.
How to cut usage
| Change | Typical effect | Cost |
|---|---|---|
| Move filters to the start | Stops later steps for non-matching runs | Minutes |
| Combine steps | One API call instead of three where the app allows | Some rework |
| Use Formatter and other free built-ins | Removes steps from the count | Minutes |
| Digest instead of per-event alerts | Replaces dozens of tasks with one | Design time |
| Batch with a scheduled job | One run processing many records | Needs a different tool or a Code step |
| Delete duplicates and unused Zaps | Immediate | Audit time |
| Reduce polling frequency | Fewer false runs (triggers are free, but downstream steps may not be) | Check |
| Move high-volume Zaps elsewhere | Compare cost on another platform | Migration effort |
Compare alternatives on your own numbers
Automation platforms count usage in different units. Zapier counts successful action steps as tasks. Make meters usage in credits, with most module runs consuming a credit, though some operations differ. n8n’s cloud plans are based on workflow executions, and the self-hosted edition has no per-execution fee but requires you to run the server, subject to its licence terms. Because the units differ, compare using your own workflow: count what each platform would charge for the same run, then multiply by volume. A multi-step workflow that is expensive per run in one unit may be cheap in another. Always check the current definition before relying on a comparison.
| Question | Zapier | Make | n8n |
|---|---|---|---|
| Unit | Task (successful action step) | Credit (generally per module run) | Execution (workflow run) on cloud |
| Cost sensitivity | Number of steps per run | Number of modules per run | Number of runs, less sensitive to steps |
| Best for | Simple, low-volume, many apps | Visual multi-step logic, medium volume | Complex logic, high volume, code, self-hosting |
| Hidden costs | Growth in volume; upgrade tiers | Credit growth in iterations | Server management if self-hosted |
The table is a framework rather than a price list. For a store whose workflows run on every order and include several steps, platforms that charge per run rather than per step can look much cheaper at volume, while for a store with few runs and many different apps, the simplicity of Zapier may be worth the premium. Our n8n, Zapier and Make service pages explain what we build on each.
A worked comparison
Take the store above, with about 9,910 tasks a month. Suppose it moves the order-to-warehouse and tracking Zaps, which use 7,500 of the tasks, to a self-hosted tool, and optimises the rest by moving filters first and replacing the review Zap with a daily digest, cutting 600 tasks. The remaining Zapier usage falls from 9,910 to about 1,800, which may fit a much smaller plan. The saving in subscription cost depends on current prices, but the pattern is clear: a few high-volume Zaps drive most of the bill. Against that saving stands the cost of hosting and maintenance, which for a self-hosted tool might be $20 to $60 a month plus time, and the migration effort of, say, 10 to 15 hours. The break-even depends on the numbers, which is why you should work them out before moving. The figures are illustrative.
Reliability matters as much as price
A cheaper bill is not worth a weaker workflow. Order-critical flows need error handling, alerts and idempotency whichever platform they run on; see idempotency and monitoring. Do not remove the monitoring step that costs a task if it is what tells you the flow has failed. Task counts should shape design, but never at the expense of safety.
What to do this week
Before changing anything, export a week of Zapier task history and sort it by Zap. Most stores find that three or four Zaps account for most of the usage. Look at each of those for filters that could move earlier, steps that could be combined and notifications that could become a daily digest. Make one change at a time, note the task count before and after, and keep the old version switched off but not deleted for a week in case something breaks. This careful routine typically trims usage meaningfully without any risk to orders, and it gives you the numbers you need to decide whether a different platform would be worthwhile.
A short audit you can do today
- Open your task history for the last 30 days and sort Zaps by usage.
- For the top three, count steps that ran per run.
- Look for filters that could move earlier and steps that could be merged.
- Delete or turn off Zaps that nobody owns.
- Find alerts that could be a digest.
- Estimate next year’s usage from your order growth plan.
- Decide whether any Zap belongs on another platform.
If you would rather have someone do this for you, our care plans include usage reviews, and our workflow automation service redesigns heavy Zaps.
Sources and further reading
Figures and rules in this guide come from the primary pages below. Platform limits and prices change, so open the page before you build on a number.
- Zapier Help Center: how task usage is measured, which steps count as tasks and which do not
- Zapier pricing, current plan limits and prices
- n8n pricing, current cloud plans and execution limits