Zapier, Make, n8n or custom code for e-commerce automation: real costs at 100 to 1,000 orders a day

Most online stores start automating with a no-code tool, and sooner or later someone asks the comparison question: n8n vs Zapier, or Make, or should a developer just write it? The honest answer depends less on features than on how each tool bills, how fast it reacts and what happens when something fails at 2am in the middle of a sale.

This guide is for store owners and operations leads who already use one of these tools, or are about to, and want to know what it will cost at 100, 300 or 1,000 orders a day. I build e-commerce automation with official APIs and custom code, and no-code is still often the right answer.

n8n vs Zapier vs Make: how each one bills

The three tools count work differently, and that decides your bill more than any feature list.

Tool	Unit you pay for	What counts	Entry paid plan (Oct 2026)
Zapier	Tasks	Each successful action step; triggers, filters, paths and formatter steps are free	Professional from $19.99/month (annual) for 750 tasks
Make	Credits	Most module actions use 1 credit; some AI features use more	About $9/month (annual) for 10,000 credits
n8n Cloud	Executions	One run of a whole workflow, whatever the number of steps	Starter €20/month (annual) for 2,500 executions

Sources: Zapier pricing, Make pricing and n8n pricing, read on 15 October 2026. Make replaced operations with credits on 27 August 2025.

The practical consequence: on Zapier and Make, every step you add to a workflow multiplies by every order. On n8n, the steps inside a workflow are free; only runs count.

Zapier also has overage rules worth reading before a busy season. With pay-per-task switched on, extra tasks are billed at 1.25 times the base rate on annual plans and 2.5 times on monthly plans, and usage stops at three times your subscription until the next cycle.

Polling, instant triggers and why minutes matter

A trigger either waits to be told (instant, using a webhook) or checks on a timer (polling). Zapier's polling interval is 15 minutes on Free, 2 minutes on Professional and 1 minute on Team. Make's is 15 minutes on Free and 1 minute on paid plans. Instant triggers fire straight away on any plan, and an n8n workflow can start from a webhook that Shopify sends the moment an order is created.

For a newsletter sign-up, two minutes does not matter. For stock it does. If the same item sells on Shopify and Amazon and your sync polls every two minutes, a busy hour gives the second channel two minutes at a time to sell something you no longer have. That is how overselling starts, and I cover the fixes in how to stop overselling across channels.

Worked example: the same workflow at 100, 300 and 1,000 orders a day

This is where n8n vs Zapier stops being a feature debate. It is an illustration, not a quote. Assume one workflow that runs per order with three actions: add a row to a spreadsheet, post to a team channel and create an invoice in your accounting tool. A 30-day month, annual billing, list prices on 15 October 2026.

Orders a day (a month)	Zapier Professional	Make	n8n Cloud
100 (3,000)	9,000 tasks: 10,000-task tier, $129/month	About 9,000 credits plus trigger runs: around the 10,000-credit entry tier	3,000 executions: Pro, €50/month
300 (9,000)	27,000 tasks: 50,000-task tier, $289/month	About 27,000 credits: roughly three times the entry tier	9,000 executions: Pro, €50/month
1,000 (30,000)	90,000 tasks: 100,000-task tier, $489/month	About 90,000 credits: roughly nine times the entry tier	30,000 executions: Business, €667/month (self-hosted)

A few things the table shows:

• Zapier scales with orders times steps. On monthly billing the same three tiers are $193.50, $433.50 and $733.50. Add a fourth action at 1,000 orders a day and you need 120,000 tasks, past the 100,000 tier.
• On Make vs Zapier alone, Make is cheaper per step, but its price at higher credit volumes comes from its pricing calculator, so I have not put a figure on it.
• n8n does not care how many steps you add, but it jumps from Pro (10,000 executions) to Business (40,000), and Business is self-hosted. The alternative is the free Community Edition on your own server.

Real stores rarely run one workflow. Refunds and fulfilment events add runs, and stock updates fire per variant change, often far more than once per order.

Where no-code breaks for stores

Cost is the visible problem. These are what cause lost orders:

• Bursts. Zapier's flood protection holds runs when 100 or more events trigger a polling Zap at once, and emails you to confirm. Replays run at one per second. A flash sale or a bulk product edit is exactly that burst.
• Rate limits. Every platform rations its API. Shopify's GraphQL Admin API refills 100 points a second on standard plans, 200 on Advanced and 1,000 on Plus. Amazon's SP-API sets limits per operation and returns HTTP 429 when you exceed them. Etsy limits per second and per day. A connector that retries blindly can hit the same wall again.
• Retries without memory. Zapier's autoreplay (paid plans), Make's error handlers and n8n's "Retry On Fail" all retry a failed step. None of them knows, on its own, that a retried "create order" must not create a second order.
• State. Sync needs memory: which Shopify variant maps to which Amazon SKU, what stock figure was last sent, which orders are already imported. Shopify's own docs say webhooks can arrive twice or out of order and delivery isn't always guaranteed, and recommend reconciliation jobs. In no-code, that memory ends up in a spreadsheet or a small data store.
• Visibility. A failed step sends an email, not a screen of every order that should have synced today and did not.

Self-hosting n8n: the licence and the real costs

n8n's Community Edition is free to self-host under its Sustainable Use License. Using it for your own business is fine, and so is paying a developer or consultant to build and maintain workflows for you. What the licence does not allow is hosting n8n as a service where your customers build their own workflows, or white-labelling it as your product.

Free software is not free to run: someone has to patch the server, back up the database, upgrade n8n and notice when the queue stops. With a developer on hand that is routine; without one, a hosted plan is safer.

When custom code wins

Custom code earns its build cost when:

1. The automation is core to selling. Stock and order sync between channels, where a missed run means an oversell or an unshipped order.
2. Volume is high and growing. A custom service's hosting bill does not rise with every step you add, while per-task pricing does.
3. The logic needs state. Mapping tables, deduplication, "only send if changed", nightly checks that both sides agree.
4. Each platform needs its own pacing. A queue that respects Shopify's points, Amazon's per-operation limits and Etsy's daily cap separately.

The shape I use is simple: webhooks land in a queue, one worker per channel processes them within that channel's limits, a database records what was sent, a reconciliation job checks for drift, and failures appear on a dashboard instead of in an inbox.

When no-code is the right answer

Plenty of store automation should stay no-code:

• glue between tools: new order to a spreadsheet, a refund to a team channel, a customer to your CRM;
• reports and alerts where a late run is annoying, not costly;
• prototypes, to prove a process before paying to build it;
• anything that lives entirely inside Shopify, where Shopify Flow is free and faster than either.

Hybrid is common and sensible: custom code for the stock and order sync, no-code for everything around it.

A quick decision guide

• Under a few thousand runs a month, nothing critical? Use whichever tool your team already knows.
• Many steps per order? That tips n8n vs Zapier towards n8n, because only runs count.
• No one to run a server? Stay on a hosted plan.
• Keeping two or more sales channels in agreement? Use custom code, or a dedicated channel app, not a general connector.
• Unsure? Count orders a day times steps per order times 30, and price it on each pricing page before you build.

What custom automation costs

A custom build is a one-off cost plus hosting in your name; a no-code subscription is a monthly cost that grows with orders. My marketplace and e-commerce app plans start at $2,500 (MVP: one marketplace through its official API and one automation, such as order export, listing updates or reports), $5,500 (Flexible: two or more channels with stock and orders kept in sync, plus a dashboard) and $9,500 (Advanced: repricing rules, AI-written listing drafts you approve, team accounts and analytics). These are starting prices, and each project gets a fixed quote after a short brief. The custom automation section shows what fits where, and the e-commerce automation guide covers what to automate first.

FAQ

Is n8n cheaper than Zapier for an online store?

Usually, once workflows have several steps. Zapier charges a task for each successful action, so cost grows with orders times steps. n8n charges per workflow run, whatever the number of steps, and its Community Edition is free to self-host for your own business. The catch is that self-hosting needs someone to maintain the server, and n8n's cloud plans jump in price between tiers. Run your own order volume through each pricing page.

What does polling mean in Zapier and Make?

A polling trigger checks for new data on a timer instead of being notified. On Zapier the interval is 15 minutes on Free, 2 minutes on Professional and 1 minute on Team; on Make it is 15 minutes on Free and 1 minute on paid plans. Instant triggers fire straight away. For stock updates across channels, even a short polling delay leaves a window where the same item can sell twice.

When should I move from Zapier to custom code?

Move when the automation keeps sales channels in agreement, when a failed run costs you an order or an oversell, or when your task count rises with every new order and step. Custom code can queue webhooks, respect each platform's rate limits, remember what it already sent and check both sides nightly. Keep no-code for glue work such as spreadsheets, alerts and CRM updates, where a late or failed run is easy to fix.

If your no-code bill or your error inbox is growing, send me a short brief: the channels you sell on, orders a day, the workflows you run now and which one hurts most. You can start from the marketplace apps MVP plan, read about marketplace integrations, or check whether Shopify Flow already covers the job.