How to Count Your Automation Runs Before Comparing Zapier, Make and n8n Pricing

Automation tools look easy to compare until you reach the pricing page. One vendor sells tasks, another sells executions, and a third sells credits. A "plan with 10,000 units" means something different in each case, so a table of entry prices tells you very little. The reliable way to compare is to count your own runs first, then price them on each platform's meter.

Step 1: Export a month of run history

Most automation platforms let you export or view a history of runs. Take one full month, not a quiet week, and include the busiest days. Save it as CSV. If you only have a screenshot or a copied table, paste it into a spreadsheet and save it as CSV so every run is one row.

Step 2: Convert it so you can filter and group

If you prefer to work with a script, a CSV to JSON converter turns each row into an object with named fields such as workflow name, status and step count. If your tool exports JSON instead, the JSON flattener makes nested step data easier to read, and the JSON key extractor lists which fields exist. To check the path your script needs, test it in the JSON path tester. When you want the grouped result back in a spreadsheet, the JSON to CSV converter reverses the step.

Step 3: Count three numbers per workflow

For each workflow, write down:

• Runs per month: how many times it started.
• Action steps per run: steps that actually do something, such as creating a record or sending a message.
• Other steps per run: filters, formatters, delays and AI steps, which some platforms count differently from actions.

Keep failed runs in a separate column. Some platforms do not bill failed actions and others do, so you want to be able to include or exclude them.

Step 4: Price the same workflow on each meter

Take one workflow and apply each platform's unit. The numbers below are illustrative assumptions, not vendor quotes. Suppose a workflow runs 2,000 times a month with three billable actions and two non-billable steps per run:

• A platform that counts successful action tasks would see about 6,000 units a month.
• A platform that counts one execution per full run would see 2,000 units.
• A platform that charges one credit per run, with extra credits for AI steps, would see 2,000 credits plus any AI-step credits.

Now repeat for your two or three biggest workflows and add the totals. Small workflows rarely change the answer. Your biggest one almost always does.

Step 5: Add growth, retries and payload size

Add a buffer for growth and retries. A fixed percentage is only a starting assumption, so re-count after the first live month. Also look at payload size: some platforms limit execution time or memory, and a large JSON body can hit those limits before the unit count does. The JSON size calculator gives a quick estimate of how heavy a typical payload is.

Step 6: Compare plans on your numbers, not theirs

With a monthly total per meter, open each vendor's current pricing page and find the smallest plan that covers it. Prices change often, so write down the date you checked. Then look at the non-price differences that matter to you: whether you can self-host, which plan includes single sign-on, and how many users each plan allows.

Where to compare the tools themselves

The ProductListo guide to the best Zapier alternatives in 2026 lists how Zapier, Make, n8n, Activepieces and IFTTT count usage, with a worked example of one workflow on several meters. Its DevTools guide covers related developer products. Makers of automation products can submit a tool to the ProductListo directory at productlisto.com/submit or read the write for us page.

Before you migrate

Whichever platform you choose, test a real payload before you rebuild a workflow. UrwaTools has a guide on testing a webhook payload with free JSON tools before you automate it, and another on keeping a readable audit log of AI agent tool calls with JSON, which also helps when you want to count runs later.