This tool originally worked with the "Order History Reports", which gave one slice of information about Amazon order history. Amazon sunset this export/reporting tool on 3/20/2023. I have since adapted this tool to work with the zipfile of data produced by that slower, cumbersome, and more comprehensive "Request My Data" tool - for details on this migration, see
this issue.
The most important change is that category information is not present in this export, meaning the tool cannot help you categorize never before seen items. For previously purchased items, by default the tool use the most common tagged category, based on an exact string match of the line item description.
Do you order a lot from Amazon? Tired of everything showing up as "Amazon" and category "Shopping" in mint.com? Then this tool is for you!
This tool takes an Amazon "Request your data" Order export and matches your Amazon order history with your Mint transactions. If it finds an exact matches it will either:
- Update the transaction description and category if there was only 1 item
- Itemize the transaction - one line-item per item in the order (via Mint transaction splits)
If the tool chooses poor categories for your transactions simply change it! The next time you run the tool it will remember your past personalized category edits and attempt to apply it to future purchases of the same item. This only works if the item names match exactly. Also, you must change all (or the majority of) all the past, tagged examples of that item for the tool to pick up the hint. Put another way: if you only change 1 past transaction and you have 10 purchases of that same item the tool will take whatever the most common category used for that item.
This tool will not retag or touch transactions that have already been tagged. Feel free to adjust categories after the fact without fear that the next run will wipe everything out. If you do want to re-tag previously tagged transactions take a look at the retag_changed option.
This tool does not save your username or password. The tool is powered by the Selenium framework to automate an instance of the Chrome/Chromium browser. When running the tool it will prompt for the username and password and then enter it into the browser for you. There are options that allow for manual user operation of the login flows for both Mint and Amazon.
This tool does not require an Amazon store card/Visa. All you need is to pay for your Amazon charges with an account that is synchronized with Mint. For example, if you alternate between 5 different credit cards to pay for purchases on your Amazon account, only the transactions from credit cards synchronized with Mint will get tagged.
Some things the tagger cannot do:
- Amazon credit card award points are not reported anywhere in the order/item reports.
- Amazon gift cards are not yet supported (see issue #59)
This project has been a passion project of mine to better understand cashflow (critical to trend analysis and budgeting).
If you have found this tool useful, please consider sponsoring me.
Please download the latest version from github's releases page
You can run the Mint Amazon Tagger via docker like so:
# Check out this git repo if not already:
git clone https://github.com/jprouty/mint-amazon-tagger.git
cd mint-amazon-tagger
# Build the image:
docker build -t mint-amazon-tagger .
# Run the container:
docker run -it --rm mint-amazon-tagger
If you're using ARM, you need to build with:
docker build --platform linux/amd64 -t mint-amazon-tagger .
-
pip3 install mint-amazon-tagger -
To get the latest from time to time, update your version:
pip3 install --upgrade mint-amazon-tagger -
Chromedriver should be fetched automatically. But if you run into issues, try this:
# Mac:
brew tap homebrew/cask
brew cask install chromedriver
# Ubuntu/Debian:
# See also: https://askubuntu.com/questions/539498/where-does-chromedriver-install-to
sudo apt-get install chromium-chromedriver
This mode will fetch your Amazon Order History for you as well as tag mint.
-
mint-amazon-tagger -
Plug in all your info into the app!
This mode will fetch your Amazon Order History for you as well as tag mint.
mint-amazon-tagger-cli --amazon_email [email protected] --mint_email [email protected]
This mode requires you to fetch your Amazon Order History manually, then the tagger automates the rest.
- Generate and download your Amazon Order History Reports.
a. Login and visit Amazon Order History Reports
b. "Request Report" for "Items", "charges and shipments", and "Refunds". Make sure the date ranges are the same.
c. Download the completed reports. Let's called them
Items.csv charges.csv Refunds.csv for this walk-through. Note that
Refunds is optional! Yay.
-
(Optional) Do a dry run! Make sure everything looks right first. Run:
mint-amazon-tagger-cli --items_csv Items.csv --charges_csv charges.csv --refunds_csv Refunds.csv --dry_run --mint_email [email protected] -
Now perform the actual updates, without
--dry_run:mint-amazon-tagger-cli --items_csv Items.csv --charges_csv charges.csv --refunds_csv Refunds.csv --mint_email [email protected] -
Sit back and relax! The run time depends on the speed of your machine, quality of internet connection, and total number of transactions. For reference, my machine did about 14k Mint transactions, finding 2k Amazon matches in under 10 minutes.
To see all options, see:
mint-amazon-tagger-cli --help
Not every bank treats Amazon purchases the same, or processes transactions as quickly as others. If you are having a low match rate (look at the terminal output after completion), then try adjusting some of the options or command line flags. To see a complete list, run mint-amazon-tagger-cli --help.
Some common options to try:
- --mint_input_include_inferred_description. This allows for more generous consideration of Mint transactions for matching. See more context here
- --mint_input_include_user_description. Similar to above; considers the current description as shown in the Mint tool (including any user edits).
- --max_days_between_payment_and_shipping. If your bank is slow at posting payments, adjusting this value up to 7 or more will increase your chance of matching. If you have a high volume of purchases, this can increase your chance of mis-tagging items.