Update on Importing Amazon Orders into Google Sheets (May 2023)

One of the most popular Tiller Community Solutions add-on workflows has long been the Import CSV Line-item workflowspecifically for Amazon purchases.

This workflow allowed users to break Amazon transactions with many disparate parts (e.g. :dog2: food, :iphone: protector, and :maple_leaf: syrup) into discrete transactions for granular categorization. To pull these line items into their spreadsheet, users simply visited Amazon’s Order History portal, downloaded the CSV, and imported that data into the TCS add-on.

In March, Amazon abruptly and bizarrely stopped providing this popular download, frustrating many users in this community. Fortunately, Amazon is required by law or industry best practices to offer users access to their personal data and has stood up a “privacy central” service to offer their customers a comprehensive record of everything they know about them… and one (small) piece of this is the order history.

Today, I updated the Import CSV Line-item workflow to read the Retail.OrderHistory.1.csv file in the “Your Orders” Privacy Central download. The data mapping isn’t exactly the same as the previous version of this tool but it is really close.

Do not miss the first bullet about Amazon Fresh purchases below… :point_down:

Some Important Notes

  • The old report did not include Amazon Fresh purchases— that data is now interleaved with the rest of your Amazon purchases— and these can be really high volume. Inexplicably, these purchases are identified in the CSV as coming from the “panda01” website. I added a filter for those of you who want to remove these items from your import (see screenshot at bottom).
  • The old report was based on a user-selected time range. The new report will contain your entire Amazon history, slowing down performance of the initial upload step considerably (it now takes several minutes for me). You may also want to use the Start Date filter to avoid importing very old transactions (e.g. beyond the beginning of your spreadsheet’s time range).
  • I believe the part of the workflow that excludes purchases that match previously-imported transactions still works (across old- and new-workflow records)— please let me know if you find this not to be the case and are seeing duplicates.
  • The new Privacy Central CSV includes information on Amazon returns, but I have not yet updated this import mapping. I hope to get to that in the coming weeks.
  • This import workflow is currently only available in Google Sheets.

I hope this helps those of you who miss the old workflow.
Randy :rocket:


Thank you! Very much appreciated.

Of course. Let me know how it goes, @lava890.
(And watch out for those “pando01” transactions.)


Randy - do you have a link to the export function? When I go to Privacy Central and request order history it acknowledges the request and says the data will be sent within 30-days.

Amazon nixed the ability to export your order history. The Privacy Central request is the only way to get it now from Amazon.

Looks very neat. Do you know if it includes Amazon.ca (Canada)? I am hoping it does since you referenced maple syrup in your post!

Edit: I just checked, and there is an option to download data via Amazon.ca, but after I confirmed, it gave me this message: We will provide your information to you as soon as we can. Usually, this should not take more than a month.

I really hope it doesn’t take more than a month.

Mine, from Amazon.com, took a couple hours to arrive.

The site says it can take days or even weeks but mine took about 24 hours. As you all receive them, let us know what your experience is in this thread.

P.S. I’m not sure on the Amazon Canada import, @galatic_incipient. Unfortunately, I’m not clever enough to hide maple-leaf emojis as confirmation that it is working. (I like your sleuthing instincts though.) Let me know how it goes for you.

Is an excel planned?

Eventually, @tjones4852, but no timeline at the moment.

Thanks for the update! I just set it up and it works great. I just have to change my existing Amazon charge to “Amazon” and the rest to whatever. I did pull in all Amazon, including Fresh, etc. which I like since I occasionally might purchase stuff in person that’s not groceries, so it’s a bit more detailed if I want.

Glad to hear it is working for you, @yossiea.

Presumably you saw the “Filter purchases after import” toggle which makes bulk changes to imported data a breeze.



I actually preferred not to filter since I had to change the original transaction from my card to “Amazon.” So I started from 2023 and then just scrolled down until I saw blank categories and put in the correct category and the CC transaction I switched to Amazon.

Thanks for coming up with a new solution – crazy that Amazon got rid of the report (again!). I’ll be testing this out in the next day!

Yes. Thank goodness for user-data protections. I thought we were going to be out of luck.

1 Like

I requested mine earlier today and got an email to download within ~2 hours. Over 1500 orders dating back to 2001! I think I probably had other orders before this that they lost track of. The very first thing they have record of me ordering? “Hamilton Beach TrueAir Plug-Mount Odor Eliminator with 3 Replacement Filters”. What the heck was going on in my life on that day…


@randy do you know when you might have amazon refunds working with the new format?
Thank you!

P.S. Does csv import tool available anywhere as open source code? (I think not, since I could not find anything, but wanted to double check).

Hoping to get to refunds in a week or so, @sagor. (Full disclosure: I haven’t reviewed the data Amazon provides so it is possible it will not be sufficient to match the previous implementation.)

The CSV tool is not open source. I’ve created a few projects in the past that might work if you want to lightly automate your workflows.

1 Like

YES!!! thank you Randy!

Downloaded and imported my amazon orders from 1997 thru 5/10/2023. How do I filter next download to start after 5/10/2023? Is Refunds still a work in progress - couldn’t get that to work.