Make Google Sheets modules: watch rows, search and update without duplicates

Our verdict

Treat the sheet a Watch New Rows trigger reads as an append-only log. Make describes it as a polling trigger that checks for new rows each time the scenario runs, and says it tracks which rows it has already processed. [2] Its documentation does not describe what happens when rows are inserted above old ones, the sheet is re-sorted or rows are deleted. Design so those things never happen to the watched tab.

For anything that should end up as one row per customer, order or email, the reliable pattern is lookup then write: Search Rows, then Update a Row or Add a Row. It costs one extra credit per record and removes most duplicate problems. [1] [4]

How Watch New Rows decides what is new

Polling triggers check for new data since the last run. Make says each run processes data the scenario has not processed before, and that later runs track changes since the previous run. [3] In the Google Sheets setup tutorial, Make says the module tracks which rows it has already processed and only fetches the ones it has not. [2]

Choose where to start applies only to the first run. Choosing All makes the module process every existing row first and then new rows. [2] [3] If you only want rows added from today, pick the option that starts from now, or you will process the whole backlog and pay a credit for each module run on each old row.

The Limit field caps how many rows Make handles in one execution cycle. [1] Make's tutorial uses 20 as an example. [2] A lower limit keeps each run short and helps you stay inside Google's quota of 60 read requests and 60 write requests per minute per user. [6] Make's rate-limit guide recommends a limit of up to 20 records per run. [9]

Header row and range rules

Set Table contains headers to Yes when row 1 holds column names; the module then leaves the header row out of the output. Row with headers takes the header range, for example A1:F1. [1] For range-based modules, if you leave the header field empty Make assumes the header is the first row of the range you specify. [1]

In Add a Row and Update a Row, Use column headers as IDs of the columns keeps your mappings working when columns move, as long as you do not rename them. [1] Rename a header and the mapping breaks quietly, so freeze header names once the scenario is live.

Search, then update or add

Search Rows returns rows matching a filter you build with AND and OR conditions, plus a sort order and a Limit. [1] The standard Search Rows result includes the row number, which Make's own examples map into modules that act on a specific row. [5] Update a Row needs that row number. [1] Search Rows (advanced) uses Google's query language but does not return a row number, so it cannot feed Update a Row. [1]

A dedupe scenario then looks like this: the trigger delivers a record with an email address; Search Rows filters the customer tab on the email column with Limit 1; a router sends the bundle to Update a Row if a row number came back, or to Add a Row if it did not. Routers and filters use no credits. [4]

Credits per poll and API quota

A trigger uses 1 credit each time it checks, even if it finds no new data. [4] Make's own scheduling tutorial shows 288 credits a day for a check every 5 minutes, 24 a day for hourly and 30 a month for daily. [8] At the default 15 minutes it is 96 checks a day, or 2,880 in a 30-day month.

Google limits the Sheets API to 300 read and 300 write requests per minute per project, and 60 of each per minute per user. Going over returns HTTP 429, and Google recommends exponential backoff. [6] Make's rate-limit guide suggests bulk modules such as Bulk Add Rows and Bulk Update Rows to process several bundles in a single request. [9] Bulk Add Rows appends rows to the bottom of the table, and Bulk Update Rows updates multiple rows in a range you set. [1]

Setup checklist

Run through this before you switch the scenario on.

• Give the watched data its own tab that only receives appended rows; do reporting, sorting and filtering on a separate tab that reads from it.
• Delete blank rows and make sure nothing upstream can create them. [1]
• Pick one unique key column, such as email or order ID, and search on it before every write. [1]
• Set Table contains headers to Yes and enter the header range. [1]
• Choose where to start deliberately on the first run; All processes the whole backlog. [2]
• Set Limit so a run handles a manageable batch, such as 20 rows. [2]
• Use the standard Search Rows, not the advanced version, when you need a row number. [1]
• Set the schedule from how fast you need the data, then multiply checks by days to budget credits. [4] [8]
• If you connect with your own Google Cloud client and leave it in Testing, expect to reauthorize every week. [7]

Sources & verification

Product details and prices can change. Check the linked provider before buying.

1. Make Apps: Google Sheets modules Accessed 30 Oct 2026
2. Make Help: Step 5, set up the trigger Accessed 30 Oct 2026
3. Make Help: Types of modules Accessed 30 Oct 2026
4. Make Help: How features use credits Accessed 30 Oct 2026
5. Make Apps: Google Sheets tips and examples Accessed 30 Oct 2026
6. Google Sheets API: Usage limits Accessed 30 Oct 2026
7. Make Apps: Google Sheets connection Accessed 30 Oct 2026
8. Make Help: Step 10, schedule your scenario Accessed 30 Oct 2026
9. Make Help: Fix rate limit errors Accessed 30 Oct 2026