Skip to content
Market note6 min read

Google Sheets preserves more of your Excel tables. Test the next row.

A converted workbook should keep working after the first screenshot. One small test shows whether its reporting relationships survived.

If your team rebuilds tables after converting Excel workbooks into Google Sheets, try the import again. Google’s August 11, 2026 release says Excel tables now become Sheets tables, and Excel pivot tables linked to table ranges retain their structure instead of becoming static grids. The useful check comes after conversion: add one row and see whether the summary includes it. Google’s release announcement.

A matching first screenshot proves very little about next month’s report. This change matters when the converted file will keep receiving data.

Market note reviewed September 8, 2026. The release details are documented; the workbook below is a fictional test we propose, not an import we have run.

What changed, and who gets it

Google lists both improvements for all Workspace customers and personal Google accounts, with no separate admin setting. Table imports were available at announcement. Linked-pivot imports began a rollout on August 11, with up to 15 days for visibility on both release tracks. That published window has passed, but we have not confirmed an individual account. Release scope and rollout.

The announcement concerns conversion into Sheets. It does not establish full compatibility for every Excel formula, macro, external connection or workbook feature, and it does not create an ongoing connection to the original Excel file.

That leaves three sensible choices:

Situation Sensible starting point The cost to consider
The team will maintain the working table in Sheets Convert a copy and test the table and pivot together One-time checking plus any remaining repair
An accountant or supplier owns the Excel master Keep their Excel file as the master; use a clearly dated converted copy only where useful Repeated imports and checking for changed columns or ranges
The workbook depends on features beyond tables and pivots Keep the established Excel process until those dependencies have been tested The risk and effort of rebuilding working logic

No new software purchase follows automatically from this release. The announced features are available in the accounts Google lists; whether conversion saves work depends on your file.

Use a workbook small enough to check by eye

Our fictional Orders table occupies A1:E4: one header row and three orders. A pivot on another sheet sums Amount by Status.

Order ID Customer Order date Status Amount
0007 North 2026-09-01 Open 100.00
0008 West 2026-09-02 Complete 200.00
0009 East 2026-09-03 Open 50.00

The independently calculated totals are Open 150, Complete 200, Grand total 350. Record those and the count of three orders before conversion. Keep order IDs as identifiers: losing the zeros in 0007 can matter even when the financial total is correct.

Now add a test order in row 5:

Order ID: TEST-01
Customer: Test only
Order date: 2026-09-08
Status: Open
Amount: 17.50

If the test row is included correctly, Open becomes 167.50 and the grand total becomes 367.50. Complete stays at 200.

Fictional workbook · expected arithmetic

The next row proves the connection

  1. Orders table before import

    A1:E4 · 3 orders

    Columns: Order ID, Customer, Order date, Status, Amount. Row 2: 0007, North, 2026-09-01, Open, 100.00. Row 3: 0008, West, 2026-09-02, Complete, 200.00. Row 4: 0009, East, 2026-09-03, Open, 50.00.

  2. Record the original pivot

    Grand total: 350.00

    Open: 150.00. Complete: 200.00. Confirm three data rows and preserve leading zeros in the order IDs.

  3. Add one test order

    A1:E5 · TEST-01

    Row 5: TEST-01, Test only, 2026-09-08, Open, 17.50. Check that the table and the pivot’s effective source include the new row.

  4. Check the expected result

    Grand total: 367.50

    Open must become 167.50; Complete remains 200.00. Both Open and the grand total rise by exactly 17.50.

    If the source still ends at row 4, a total of 350.00 misses the test row.

  5. Remove the test order

    Grand total returns to 350.00

    Open returns to 150.00; Complete stays 200.00. Confirm that the test row is gone and the original three orders remain.

A matching screenshot after import is not enough. The report must include the next row and return to its starting totals when that row is removed.

Sources: Google’s import release · Sheets tables · Sheets pivots

Proposed import check using fictional orders. The expected totals are calculated; no Sheets import has been observed for this article.

Google documents that named table references update as table data changes. It also documents that pivot tables refresh when cells in their source change. The test is to establish whether the imported pivot and the newly extended table still have the relationship you need. Sheets table references, Sheets pivot behavior.

Run these checks on your actual import

Work on a copy of a representative Excel workbook, with the original available beside it.

  1. Confirm the original. Record its data-row count and the totals you rely on. Note the table name, pivot source, filters and any deliberately excluded rows.
  2. Convert using your normal route. Check that you are inspecting the converted Google Sheet, not merely viewing the Excel original.
  3. Inspect the imported table. Check headers, boundaries and column types. Open one date, an amount, a blank and an identifier with leading zeros.
  4. Inspect the pivot. Check its source, grouping and filters. Does Amount summarize by sum, or has the import left it counting records?
  5. Add the test row inside the table. Inspect whether the table includes row 5 and whether the pivot’s effective source includes it.
  6. Check the exact change. In the fictional example, only Open and the grand total should rise by 17.50.
  7. Remove the test row. Confirm that the count and totals return to their starting values.

A total that initially matches can conceal a static output. A total that changes can still conceal a duplicated record. That is why the row count, category totals and grand total all belong in the check.

Diagnose the mismatch before repairing it

Symptom Inspect next Avoid this shortcut
Grand total remains 350 after the test row Is row 5 inside the table and the pivot’s effective source? Typing 367.50 into the report
Grand total changes but Open does not Status spelling, filters and blank categories Renaming a displayed label without checking the source
Amounts appear as counts Pivot value settings and whether imported amounts are numeric Multiplying the count to imitate the old total
Totals match but an order is missing Row count, duplicate IDs and filters Accepting the grand total as the only check
Dates or IDs changed meaning Column types, underlying values and relevant file settings Repairing only the visible formatting

These are proposed diagnostic steps, not observed defects in the release. The spreadsheet validation guide includes a practice workbook and explains how to check values before they enter a report.

Test the second import too

A monthly supplier workbook creates a different job from a one-time migration. Make a second copy with an extra column and a later month’s rows, then run the same import route. Establish whether the team must reapply filters, types or checks each time.

Illustrative effort calculation: if rebuilding a report currently takes 12 minutes and the new route takes five minutes to inspect and repair, the proposed saving is seven minutes per import. Across six workbooks a month, that would release 42 minutes. If the new route still takes 12 minutes after checking, it has not reduced preparation time. Those figures are assumptions to replace with your own observations.

Agree which file owns the next update. A converted Sheet and an Excel master can both look authoritative while disagreeing. Label the derived report with its source filename, owner and “as of” date, and keep the approved import steps in the SOP template.

This release deserves a retest if rebuilding tables is part of your routine. Keep the conversion if the next row, the next reporting period and the next colleague all work with less repair. Use the software pilot scorecard to preserve that evidence.

Back to the blog