Build your first Make.com business automation.
Move approved sample requests from an intake sheet to a work queue. This small Make.com project teaches a trigger, a filter, and field mapping with an output you can inspect. It is a separate practice exercise to pair with the one-hour Make.com automation course.
You need Make and Google Sheets accounts. Use a new test spreadsheet with fictional records. The guide is free; scenario runs consume your account’s allowance. Keep the schedule off while testing.
1. Prepare two tabs
Create a spreadsheet with tabs named Intake and Work queue. Give both the same first-row headers: Request ID, Task, and Approved. In Intake, add these rows directly below the headers, without blank rows:
| Request ID | Task | Approved |
|---|---|---|
| TEST-001 | Draft a project brief | yes |
| TEST-002 | Review an idea | no |
Leave Work queue empty apart from its headers. No customer details, emails, or external notifications are needed for this exercise.
2. Watch the intake rows
Create a scenario and select Google Sheets → Watch New Rows. Connect the Google account that owns the test sheet. Select the spreadsheet and Intake tab, enable headers, use header range A1:C1, and set the limit to 2.
For this new test sheet, select All when choosing where to start. That includes the two sample records. Save the module and the scenario. Make’s trigger setup guide explains the starting position and how polling remembers processed rows.
3. Map the destination
Add Google Sheets → Add a Row and choose Work queue as its destination. Map each destination column to the corresponding output field from Watch New Rows. Select Raw for the value input option so values are stored as supplied. Choose Insert rows if that option is shown.
The Google Sheets module reference documents the fields and notes that a blank row can stop Watch New Rows from processing later rows.
4. Pass only approved requests
Add a filter on the connection between the modules. Label it Approved requests. Use the trigger’s Approved field, the text equality operator, and the value yes. The destination should receive only rows that satisfy this condition. See Make’s filter guide for the current controls.
5. Run and verify
Save, select Run once, and inspect the module results and the Work queue tab. These are expected results to verify in your own account, not a claim that this scenario has been run on your behalf:
- TEST-001 appears once in Work queue with the same task and approval value.
- TEST-002 does not appear because its Approved value is no.
- A second run with no newly appended intake rows should add no rows.
- Append TEST-003 with a new task and Approved set to yes; the next run should add that row.
Understand the limits before scheduling
This is an append-only exercise. Watch New Rows is not an approval-update trigger: changing an already processed row from no to yes will not make it a new row. For a real approval queue, design an update-aware workflow or search for pending records on a schedule.
The trigger cursor also does not guarantee exactly-once delivery. Resetting the starting position, copying records, or retrying a partially completed run can produce duplicates. Before using real requests, decide how a stable request ID will prevent repeated processing and how failed runs will be reviewed.
If nothing arrives, check the selected tabs, starting position, header mapping, filter value, and blank rows. If every row arrives, check that the filter is on the route to Add a Row. After the expected results pass, choose a schedule appropriate for your account’s limits and monitor its run history.
Add AI only when the record flow works
A later project could draft a summary for each accepted request. First establish that the right input reaches the right destination. The HeyGen course companion shows how provider IDs and API limits can interrupt a larger Make workflow. For API fundamentals, use the local request exercise.