Toolkit
The 6-Question Automation Check
Find the tasks worth automating, and catch the ones that will quietly fail, send the wrong thing, or waste your weekend
Get it on GumroadDecide which task to automate before you spend a weekend building the wrong one.
You already use Zapier, Make, n8n or an AI agent. The question is not can this task be automated. It is should it be, and what has to stay human.
We read about 130 first-hand accounts from small-business owners and freelancers in the Zapier, Make and n8n forums, Reddit, Indie Hackers and X. The same three things kept happening:
- Silent stops. A freelancer's lead-capture workflow stopped working and went unnoticed for three days; about 40 leads were lost and the contract was terminated. At least seven people described a failure that ran unnoticed for three days to three weeks.
- Confident wrong sends. One operator's automation ran clean for six days, then emailed the wrong person on day seven, with nothing flagging the mistake. A double charge from a retried webhook. A lost deal from robotic client emails.
- Builds that should never have started. "Just spent 3 hours rebuilding a Zap that should've taken 10 minutes." "I spent 6 hours automating a 5 minute task." Three to six hours per episode, sometimes weeks.
Free checklists and the $19 to $27 ROI toolkits score time saved. None we found will refuse to say GO when a wrong send or a silent stop would hurt you and nobody checks first. This one does.
How it works. One task. Four scores (value, repeatability, input consistency, safety). Two yes/no questions. The spreadsheet returns GO, NO-GO, INCOMPLETE or CONTRADICTION by formula, and it will not return GO on a missing score or on "safe on paper but harmful if it fails and no human checks". If GO, you write down the human checks that stay in place before and after it runs.
What's included
- Guide (PDF): the method and how to read the four decisions, in one sitting
- Qualification worksheet (Excel with the decision logic in cells, plus a fillable Markdown version)
- Human review checklist (PDF): before-run and after-run checks
- Two worked examples (PDF): one task that scores GO with the checks that made it safe, and one that scores NO-GO built from the lead-capture and looping-API incidents above
What it is not. It does not build the automation, choose a vendor, or promise an ROI. It is one decision, done properly, before you build.
Made by 4NLab, a boutique technology partner. We're a team of developers, designers and strategists with many decades of combined experience helping startups launch and enterprises modernize, and we build AI and automation tools for enterprise clients. We have seen first-hand what breaks when a task is automated without the right checks. This toolkit is that judgment, written down. Research collected September 2026; every quote traceable to a public first-hand post.