Step 2: Prepare the file · MS Gong Ride Setup

Goal: turn the raw export into a file containing only the orders you have not already imported, with the one formatting bug fixed.

This is the step that goes wrong. Take your time with it.

1. Strip the trailing apostrophes. Not optional.

Funraisin appends a single quote character to the end of every Product ID and Variant ID value. So a variant that should read 47221186068636 comes out as 47221186068636'.

Shopify rejects every one of those rows with the error "expected String to be a id". If your import fails on every line, this is why, one hundred per cent of the time.

Note

Read this before you open the file in Excel. Variant IDs are 14-digit numbers. While the apostrophe is attached, Excel treats them as text and they are safe. The moment you remove it, Excel can decide the cell is a number and redisplay it as 4.70969E+13. Save as CSV at that point and the shortened value is what gets written to the file. It matches no product in Shopify, so the order imports with the item silently missing. This is the same failure that left the socks unfulfilled for months, and nothing on screen warns you it has happened. Google Sheets does not do this, so it is the safer choice if you have it.

Option A, Google Sheets (recommended):

File, then Import, then upload the CSV, and choose Replace spreadsheet.
Select the Product ID column. Edit, then Find and replace. Find ', leave Replace empty, Replace all.
Repeat for the Variant ID column.
File, then Download, then Comma-separated values.

Option B, Excel:

Open Excel first, with no file. Go to Data, then From Text/CSV, and pick your file.
In the preview window choose Transform Data.
Click the Product ID column header, then Transform, then Data Type, and set it to Text. Do the same for Variant ID. This is the step that protects the numbers. Do not skip it.
Choose Close & Load.
Now do Find and Replace on each of those two columns: find ', leave Replace empty, Replace All.
Save As, and choose CSV UTF-8.

Then check it worked, whichever option you used. Open the saved CSV and look at any Variant ID. It must read as a full 14-digit number, for example 47221186068636. If you see anything containing E+, the file is corrupted. Do not import it. Start again with Option A.

Note

Do not fix this by editing the EZ Importer mapping. That mapping is shared with the Walk and Mega events. Changing it to suit one event breaks the others. Always clean the file instead.

2. Remove rows you have already imported

EZ Importer has duplicate checking switched off. It will happily import the same order a second time and you will end up with two of everything, two ticket emails to the buyer, and two items sent to the fulfiller.

Keep a simple record of what you have already brought in. The easiest method:

Keep a running file of imported orders, listing the Name and Variant ID of every row you have imported.
Each time you pull a fresh export, remove any row whose Name and Variant ID pair already appears in that record.
What is left is genuinely new.
After a successful import, append the new rows to your record.

Matching on Name alone is not enough. A buyer can come back and order again under a new Name, and a single order legitimately has several rows.

3. Take out the test orders

Staff test orders should never reach a fulfiller. They usually look like a real name with a plus sign in the email address, for example firstname.lastname+9@msplus.org.au. On Gong Ride the known test orders are GONGS149194, GONGS149197 and GONGS149149.

Also drop any row whose Financial Status is not paid, and any row with a blank Variant ID. A blank variant means the product is not linked yet, so go back to The Critical Link and fix the mapping before importing that item.

4. Keep each buyer's items together

Import one order per buyer containing all of their line items. Do not split merchandise, socks and tickets into separate imports.

Shopify splits the order across fulfillers by itself, based on where each item's stock sits. Splitting it yourself creates several orders for one buyer, which means several payments to reconcile and a refund mess to untangle later. That exact situation cost a great deal of time on Gong Ride in July 2026 and it is entirely avoidable.

Worked example, Gong Ride

A July 2026 export returned 1317 rows. After removing already-imported rows, test orders and unpaid rows, 64 genuinely new orders remained, made up of 20 merchandise, 50 ticket and 12 sock line items. Those 64 went in as 64 combined orders, one per buyer, with zero errors.