Reorder Excel Columns to Match Your Database Table Schema

Match a Postgres CREATE TABLE for positional COPY

Your table is users(id, email, full_name, created_at) but the export columns are scrambled. Reorder to the schema order so COPY (no column list) binds correctly.

Table: CREATE TABLE users (id INT, email TEXT, full_name TEXT, created_at DATE);

Export (.csv) before:
full_name,created_at,id,email
Sue Lee,2026-01-04,1042,sue@x.com

Reorder to: id, email, full_name, created_at

After (.reordered.csv):
id,email,full_name,created_at
1042,sue@x.com,Sue Lee,2026-01-04

Import:
COPY users FROM 'users.reordered.csv' WITH (FORMAT csv, HEADER true);

SQLite .import binds by position

SQLite's .import maps the first CSV column to the first table column. A wrong order quietly loads the wrong fields; reorder first.

Table: CREATE TABLE orders (order_id, customer, amount);

Before: customer,amount,order_id
After reorder: order_id,customer,amount

sqlite> .mode csv
sqlite> .import orders.reordered.csv orders

Verify: SELECT * FROM orders LIMIT 3;  -- amounts now in amount, not customer

Silent wrong-field load caught by spot-check

Both state and city are text, so a swapped order imports without error — but the data is wrong. Reordering to schema order prevents it; the post-import SELECT confirms.

Table: addresses(id, city, state, zip)

Wrong export order: id, state, city, zip  → COPY succeeds, no error
SELECT id, city, state FROM addresses LIMIT 1;
  1 | WY | Leeds   <-- city and state swapped!

Fix: reorder export to id, city, state, zip, then re-import.

Extra column the table doesn't have

The export includes a notes column the table lacks. Reorder the real columns to the front, push notes to the end, then remove it before a positional import.

Table: products(sku, name, price)
Export: name,notes,price,sku

Step 1 reorder to: sku, name, price, notes
Step 2 csv-column-remover drops 'notes' ->
  sku,name,price
Step 3 COPY products FROM 'products.csv' CSV HEADER;

Excel export aligned and re-saved as CSV for COPY

Your data is an .xlsx. Reorder here (you get an .xlsx back), then save that as CSV in Excel for a CSV-only importer like COPY.

Upload sales.xlsx -> reorder to id, region, total -> download sales.reordered.xlsx

In Excel: File > Save As > CSV UTF-8 -> sales.reordered.csv

Postgres:
COPY sales FROM 'sales.reordered.csv' WITH (FORMAT csv, HEADER true);

When two out-of-order columns share a type (both text, both numeric), a positional import succeeds with no error and silently misplaces the data. There is no warning at load time. Reorder to schema order first and always run SELECT * FROM table LIMIT 5; to confirm placement.

Reorder never renames. For name-binding importers (pgAdmin wizard, Workbench), a header of email_addr will not map to a email column no matter how it is ordered. Rename with csv-header-rename first. For position-binding importers, header text is ignored entirely.

This tool keeps every column. A positional COPY then fails with a column-count mismatch, or maps extras into the wrong fields. Remove surplus columns with csv-column-remover, or give COPY/INSERT an explicit column list so extras are ignored.

The engine flattens the sheet to values before reordering, so a = formula imports as its last-calculated result. That is usually what you want for a database load, but recalc the workbook in Excel first if a formula was stale.

A multi-tab workbook is reduced to its first sheet. If your import data lives on a later tab, move it to the front in Excel before uploading, or the wrong sheet's columns will be reordered and downloaded.

The tool keeps empty cells empty — it does not convert blanks to NULL. How an empty field loads depends on your importer's NULL setting (e.g. COPY ... NULL ''). Configure that in the importer; reordering does not change it.

Column Reorder is Pro-gated. The free tier shows an upgrade prompt instead of processing. Pro raises limits to 50 MB / 100,000 rows; large database extracts usually need Pro + Media (200 MB / 500,000 rows) or Developer (500 MB, unlimited rows).

Reorder copies cell text verbatim; it does not normalise date or number formats. A 04/01/2026 that your database reads as US MM/DD versus UK DD/MM is a format issue, not an order issue. Standardise first with excel-date-standardizer.

Row 1 is always read as the header for naming the reorder list, and HEADER true in your import skips it. If your file has no header row, the first data row becomes the header and one record is lost. Add a header row or set HEADER false accordingly.