When you say “I want to automate this,” what you usually mean is: “I keep doing the same small steps, and I’d like a reliable helper.” On Windows, Google’s tools can be that helper—but only if you pick the right one for the job.
{{img:hero}}
Think of this guide like a set of labeled boxes in a workshop: once you put the work in the right box, the rest gets simpler.
Below is a decision tree you can follow, with beginner examples and the “why” in plain language.
Step 0: Describe the job in one sentence (your “sticky note”)
Before tools, write one sentence that starts with “When…” and ends with “so that…”
- When a customer fills a form, send them a confirmation email so that they know it worked.
- When I add a row to a sheet, create a calendar event so that I don’t forget deadlines.
- When an email arrives with an invoice PDF, save it to Drive so that it’s organized.
This sentence becomes your requirements. If a tool can’t clearly do that sentence, it’s the wrong tool (or you’ll need a connector).
Decision Tree: What “starts” the automation?
Most automations are triggered by one of three starters. Pick the closest match.
- A person submits something (a form, a request, a checklist)
- Time happens (daily, weekly, at 9am, first day of month)
- New data appears (a new row, a changed value, a new email)
Now follow the branch below that matches your starter.
Branch A: If a person submits something → start with Google Forms + Sheets
{{img:forms-sheets}}
Analogy: Forms is the “front door” where requests come in. Sheets is the “inbox tray” where they land in a structured way.
- Choose this when you need consistent input (names, dates, options) from people.
- Avoid this when the input is unstructured (long back-and-forth, lots of exceptions). That’s more of an email/thread situation.
Beginner example: “New equipment request” form that feeds a Sheet. From there, you can sort by priority, assign owners, and track status.
Next decision: what should happen after the submission?
- If it’s just collecting + reporting → Forms + Sheets may be enough.
- If it must send messages, create events, or enforce rules → you’ll likely add Apps Script (or a lightweight add-on/connector).
Keep your first version boring: collect the data reliably before you try to do five automatic actions.
Branch B: If time happens → start with Google Calendar (or a scheduled Apps Script)
{{img:calendar-clock}}
Analogy: Calendar is your “bell tower.” It rings on schedule, whether you remember or not.
- Choose Calendar when the automation is really a reminder, a recurring task, or a predictable cadence.
- Choose a scheduled script when you need to do work automatically at a time (generate a report, clean up data, email a digest).
Beginner example (Calendar): A recurring event called “Send weekly status update” with a saved email template link in the description.
Beginner example (Scheduled script): Every Monday at 8:30am, email yourself (or your team) a short summary pulled from a Google Sheet.
If your goal is “make sure humans do a thing,” Calendar is often enough. If your goal is “the computer does the thing,” that’s where a scheduled script fits.
Branch C: If new data appears → decide where the data lives (Gmail, Sheets, or Drive)
“New data” is vague on purpose. The key is where it shows up first.
- New email arrives → start in Gmail (filters/labels first; script second)
- New row/value changes → start in Sheets (rules, validation, then script)
- New file appears → start in Drive (folders, naming, then script)
Pick one “source of truth.” If you try to treat email, a sheet, and a doc as equal sources, you’ll constantly wonder which one is correct.
Now decide: Do you need “rules,” “messages,” or “systems talk”?
{{img:script-gears}}
This is the part people skip—and then they accidentally choose a tool that can’t finish the job.
- If you only need rules inside a sheet (required fields, allowed values, basic calculations) → use Sheets features first (data validation, protected ranges, formulas).
- If you need to send or route messages (confirmation emails, alerts, reminders) → Gmail + templates can help, but automation often points to Apps Script.
- If you need systems to talk (create Calendar events from rows, write to multiple sheets, generate docs, move files) → Apps Script is the “glue.”
Analogy: Sheets rules are like lines painted on a road. Apps Script is like traffic lights that can react to what’s happening.
A practical mini decision tree (quick picks)
- If you need structured input from people, then Forms → Sheets.
- If you need a repeating cadence, then Calendar reminders or a timed Apps Script.
- If your “trigger” is an email, then Gmail filters/labels first; add Script if you must extract data or save attachments.
- If you must create/update Google items automatically (events, docs, folders), then Apps Script is usually the cleanest Google-native option.
- If the job crosses lots of apps and needs little code, then consider a connector later—but only after you’ve defined the one-sentence job.
On Windows, the “best” choice is usually the one that reduces manual steps without creating a new maintenance hobby.
Common beginner pitfalls (and the calm fixes)
- Pitfall: Automating before the process is stable. Fix: run it manually for a week; write down the exceptions you hit.
- Pitfall: Using a Sheet like a database with no structure. Fix: add headers, freeze row 1, enforce data validation, and decide what each column means.
- Pitfall: Trying to read meaning from email text. Fix: use a Form, or at least standardize subjects with a template.
- Pitfall: Too many triggers (time + edits + email) firing unpredictably. Fix: pick one primary trigger and log the actions somewhere (even a “Logs” sheet).
Small automations succeed when they’re boring, predictable, and easy to explain to your future self.
Takeaway: Pick the “starter,” then add only one layer of complexity
Use the decision tree in this order: starter (person/time/data) → source of truth (Forms/Sheets/Gmail/Drive) → actions (rules/messages/systems talk).
If you’re new, aim for one automation that removes one annoying step. Once that works reliably, you’ll know exactly what to automate next.