The workbench · Version 1.0
Automation workflow usage and cost calculator
A task, a credit and an execution are different units. Count the work first, then apply an actual plan quote. No vendor is declared cheapest from a headline allowance.
Zapier: records × standard chargeable actions. Make: trigger checks/runs + records × (bundles × work modules + chargeable helpers). n8n: one event-driven workflow execution per record. A user-entered percentage adds uniform billable work, with totals rounded up.
Model a month
Examples:
Three reproducible examples
| Case | Inputs | Modeled monthly units |
|---|---|---|
| Form, three actions | 400 records; 3 actions; 400 trigger runs; 1 bundle × 3 modules; no helpers or retries. | 1,200 Zapier tasks; 1,600 Make credits; 400 n8n executions. |
| Order, three item rows | 400 orders; 3 row-writing actions; 400 trigger runs; 3 bundles × 1 module plus 1 chargeable iterator; no retries. | 1,200 tasks; 2,000 credits; 400 executions. Assumes the Zapier loop utility itself is zero-task and the Make iterator counts. |
| Polling plus extra work | 100 records; 2 actions; 2,880 Make checks; 1 bundle × 2 modules; 10% extra work. | 220 tasks; 3,388 credits; 110 executions. n8n assumes event-driven starts, not the same polling schedule. |
What to watch out for
Zapier’s built-in tools can be zero-task, while AI, MCP and extended code may consume more than one task. Make exempts some control modules but counts iterators/aggregators; advanced steps can cost more. n8n has concurrency, timeout and retention limits, plus separate assistant or external-service costs. The n8n illustration assumes one execution per incoming record; scheduled workflows or batching require a different input interpretation. Make’s trigger input counts checks or runs, including webhook starts. A check returning several records is still one trigger operation. These are normalization assumptions, not claims that all three connectors behave identically.
The retry input is a budgeting scenario, not an observed failure rate. It adds work uniformly, including trigger checks/runs in the Make total. Real retry charging differs by step and vendor. The optional cost calculation is included price + excess units × entered linear rate. It does not model tier changes, overage opt-in/caps, nonlinear credits, tax, promotions, currencies, external model bills or failed-action exceptions. Above an allowance with no extra-unit quote, the total stays unknown.
Use the model in a decision
Map one actual connector path before setting counts. Read Make vs Zapier’s billing comparison and the alternatives decision. For a concrete multi-bundle case, see one Shopify row per line item. Keeping a working automation is a valid nonpurchase option.
Current billing definitions
- Zapier pricing and task reference.
- Make plans and credits; operations documentation.
- n8n plans and execution definition.
Sources read November 7, 2026. Plan prices are intentionally user-entered rather than a stale automatic “cheapest” verdict. Inputs stay in memory; the calculator makes no network requests and saves no inputs.
Part of Automation & CRM. The calculation is useful without buying or signing up.
Citation and revision
The Guide Index, “Automation workflow usage and cost calculator”, version 1.0, November 7, 2026. Canonical: https://theguideindex.com/tools/automation-cost/. Original calculations and synthetic examples, not a vendor test or certification.