I cancelled Zapier, Make, and n8n

Hi Reader,

I keep seeing the same thing with the firms I talk to, and it's a little painful to watch.

Companies are spending hundreds, sometimes thousands, of dollars a month on Zapier subscriptions they don't need.

Nobody set out to. It just stacks over time.

One Zap here. A "premium" tier there. Per-task fees that creep up every quarter. And the thought of auditing it all, figuring out what each Zap even does anymore, feels too painful. So the bill keeps running.

Someone I was talking to recently killed an $800-a-month Zapier bill in an afternoon.

All it took was some focus and Claude Code.

I wasn't spending $800 a month. But when I saw that, I went and cancelled my own no-code connector tools too. Make. Zapier. n8n.

And here's the part that surprised me: I still run all of it. Lead scoring, data cleaning and enriching, notifications, inbox clearing. Every automation those tools were running still runs. It just costs me almost nothing now.

Why the math changed

The real value in Zapier was connecting and automating things without needing a developer. That was worth real money for a long time.

Except now we all have those developer skills. An AI can write, test, and maintain a small script that moves data between two apps— the exact job most Zaps do. What used to require hiring someone now takes an afternoon of asking questions.

For engineering and construction firms this shows up everywhere: the Zap that copies proposal data into the CRM, the one that files project emails, the one that pings a channel when an RFP lands. Most of them are simple two-app moves. Those are the ones to rebuild.

When keeping it makes sense

A few workflows do still belong on the platform, and it's worth knowing which ones before you start cancelling.

Somebody has to fix it when an app changes. This is the big one. When a connected app changes its API, Zapier fixes the connector for everyone. A script needs someone to notice it broke, figure out why, and patch it. If the person who built it leaves (or was the one AI-curious person on the team), that script is now a mystery. If a workflow is business-critical and you don't have someone watching it, the subscription is cheaper than the surprise.

Oversight, if you're on the tier that has it. Zapier's enterprise plan gives you audit logs, permissions, version history, and one place an admin can see every automation. A folder of scripts gives you none of that. Worth it for workflows touching client data or financial systems. On the $30-a-month plan, you're paying for the connectors, and the governance features live a few tiers up.

The plumbing you'd have to build. Retries, error alerts, run history, keeping logins refreshed— the platform handles all of it. A script has to do that work too, or it fails silently at 3 AM. AI can write that plumbing, but it's part of the honest cost of rebuilding.

With the clients I work with, though, these usually turn out to be the exception. Most of them are spending hundreds, if not thousands, of credits moving data from one place to another. Simple things a script can do. AI is ready to build that— or a builder on your team is.

Run the audit

Here's the prompt. Run it with browser use enabled in whatever AI system you use:

"I pay for [Zapier/Make/etc] and want to know which of my automations I could rebuild as simple scripts and stop paying for.

Look at my workflows— if I've pasted screenshots, work from those. If you can use my browser, I'm logged into Zapier: open my Zap list, click into each one, and read the steps yourself.

For each workflow, tell me:

Is it simple (just moving data between two apps) or complex (branching logic, many steps, human approvals)?
Good candidate to replace with a small script, or leave it on the platform?
What rebuilding it would take— which apps/APIs are involved, rough effort
Then rank them, easiest to hardest to replace

If anything's unclear from what you can see, ask me before you answer."

Try it. Even if you replace half your Zaps, that's real money back every month— and a clearer picture of what your automations actually do.

This is the same thinking I wrote about in The Tool Works. People Just Run Out of Ideas.— the capability is sitting there; the gap is knowing where to point it.

Let's talk about yours

I work with engineering and construction firms on exactly this— finding the workflows worth rebuilding, and the ones worth leaving alone. One hour a week. Your actual tools and subscriptions. Let's connect.

Hit reply and tell me what the audit finds. I'd love to hear the number.

Keep building,