Debugging a Broken Automation: My Step-by-Step Process

You built an automation to save time. Instead, it’s eating more of it. You dread the process of “debugging” a workflow, trying to find and fix the issue.

Maybe it’s a workflow that stopped firing. Maybe leads are slipping through the cracks because a “Zap” broke three days ago, and nobody noticed. Or maybe the whole thing “works” but spits out the wrong data half the time.

You’re not alone. And honestly? This is one of the most common frustrations I hear from business owners who’ve dipped their toes into automation.

The tool was supposed to free up your time.

Now it’s just another thing on your plate.

The Real Cost of Broken Automations

Here’s where it gets uncomfortable.

UK businesses that use automation effectively are seeing an average ROI of 312% within 18 months. Some SMEs report returns of 200% to 400% in the first year alone.

That means every week your automation sits broken or half-working, you’re not just wasting time. You’re leaving serious money on the table.

And the time cost?

Businesses that get automation right save 15 to 40 hours per week per employee.

That’s not a typo per employee.

So when your automation breaks and you spend three hours trying to fix it yourself, that’s not a minor inconvenience. That’s a strategic problem compounding week after week.

Something has to change.

There Is a Better Way: Debugging An Automation

What if you had a repeatable process for diagnosing and fixing broken automations?

A simple framework you could follow every single time, without needing to be technical or hiring a developer on the spot.

That’s exactly what I’m going to walk you through.

What Life Looks Like on the Other Side

Imagine this. It’s 4 pm on a Thursday, and you’re closing your laptop.

Your client onboarding runs itself. Invoices go out without you touching them. Follow-ups happen on time, every time. And when something does go wrong, you know exactly where to look and what to do.

You pick the kids up from school.

You’re actually present at dinner.

Not half-there, thinking about that workflow you need to check.

Your business grows. Your hours don’t. That’s not a fantasy. That’s what happens when your automations are solid, and you know how to keep them that way.

My Step-by-Step Debugging Process

This is the exact process I use when an automation breaks. Whether it’s mine or a client’s, I follow these same steps every time.

It works for any tool. Make, Zapier, n8n, whatever you’re using. The logic is the same.

And the best part? Once you learn this, you can fix most issues in under 30 minutes.

Step 1: Don’t Panic. Check the Trigger First.

Every automation starts with a trigger. A form submission, a new row in a spreadsheet, a payment received. Whatever kicks the whole thing off.

Nine times out of ten, when an automation “breaks,” the trigger is the problem.

Here’s what to check:

Is the trigger app still connected?

Tokens expire. Passwords change. Connections drop.

Did the trigger event actually happen?

Sometimes the issue isn’t the automation. It’s that the form wasn’t submitted or the spreadsheet wasn’t updated.

Has the trigger format changed?

If someone edits a form field or renames a column, the automation can’t find what it’s looking for.

Start with the trigger. You’ll solve roughly half your problems at this stage.

Step 2: Walk Through the Steps One by One

If the trigger is fine, the issue is somewhere in the middle.

Open up your automation and look at the execution history or run log.

Most tools show you exactly where things went wrong. There’ll be a red flag, an error message, or a step that simply didn’t fire.

Here’s what to look for:

Failed steps:

Click on them. Read the error message. It’s usually more helpful than you’d think.

Empty fields:

A step might be pulling data from a previous step that’s now returning blank. This often happens when field names change.

Timing issues:

Some steps need a delay. If step 3 tries to pull data before step 2 has finished writing it, you’ll get errors.

Don’t try to fix everything at once. Find the first broken step. Fix that. Then run it again.

Step 3: Test with Real Data (Not Sample Data)

This catches people out all the time. Including me.

I’m going off track. But this is why I get any automations I build handed over to a client asap. Then we can test an automation with real-world data.

When you first build an automation, you test it with a sample record. That sample might have every field filled in perfectly. But real data is messy.

A client leaves the phone number field blank. Someone enters their email in the name field. A date comes through in the wrong format.

To properly debug, run your automation with the actual data that caused the problem.

Most tools let you replay a specific execution. Use that feature.

You’ll often find the issue is a data problem, not a tool problem.

Step 4: Check Your Connections and Permissions

This one’s boring but important.

Every automation connects to external tools.

Your CRM, your email platform, your invoicing software. Those connections rely on API keys, OAuth tokens, and permissions.

Here’s what goes wrong:

Expired tokens:

Most connections need re-authenticating every few months. Some every few weeks.

Changed permissions:

If someone on your team changes the access settings on a Google Sheet or a shared folder, the automation loses access.

Updated apps:

When a tool you use pushes an update, it can change the way its API works. Your automation might need adjusting.

Go through each connected app and make sure the link is still live. Reconnect anything that looks off.

Step 5: Simplify and Rebuild the Broken Section

If you’ve checked everything above and you’re still stuck, it’s time to simplify.

Take the broken section of your automation and rebuild it from scratch. But simpler.

Here’s how:

Delete the broken steps.

Recreate them one at a time.

Test after each step before adding the next.

This sounds tedious. But it’s faster than staring at a complex workflow trying to spot what’s wrong.

Often, rebuilding reveals the issue immediately. Maybe you accidentally mapped the wrong field. Maybe a filter condition was set to “and” when it should have been “or.”

Small mistakes cause big headaches.

Rebuilding step by step highlights these far quicker than trying to review and inspect all the moving parts of a workflow

Step 6: Add a Safety Net for Next Time

Once it’s fixed, don’t just move on. Add monitoring so you catch problems early next time.

A few things that take five minutes but save hours:

Error notifications:

Set your automation tool to email or Slack you when a step fails. Every major platform supports this.

Weekly check-ins:

Block 15 minutes once a week to glance at your automation logs. Catch small issues before they become big ones.

Documentation:

Yes, I know this is BORING!

Write a one-line description of what each automation does and when it was last updated. In the future, you will be grateful you did this.

This is the step most people skip. It’s also the step that separates automations that run for years from ones that break every month.

Conclusion: Debugging A Broken Automation

Broken automations don’t mean automation isn’t for you.

They mean you need a process.

I know the word debugging sounds daunting. It comes from the software development world. It means “fault finding and fixing.

Follow these six steps, and you’ll fix most issues yourself in less time than it takes to brief a developer. And once your automations are running properly, you get that time back. Every single week.

Ready to Get Your Automations Working Properly?

If you’ve got automations that aren’t delivering or you’re not sure where to start, our Automation Health Check is designed for exactly this.

It’s a 2-hour session where we map out your biggest time-sinks, identify what to automate first, and give you a clear plan you can act on. All for £500.

Book your Automation Health Check here.

Alternatively, you can book a free 30-minute strategy call to help you get up and running using AI and automation.

Debugging an Automation: FAQs