How I used ChatGPT to analyze my massive Kindle library – and the mysteries it revealed

Have you ever wanted a complete list of all the books you bought on Amazon? Do you want to know which are physical books, which are Kindle ebooks, and which are Audible books? Do you want to slice and dice your Amazon book collection by genre, date purchased, or author?

My wife did. She has thousands of Kindle books and probably as many Audible books. She has been an avid reader since she was a little girl, and her book collection, both physical and virtual, is a pride and joy.

Also: How to transform your old, obsolete Kindle into the ultimate open-source reader

She’s also very analytical. A few months ago — back when we were saving her Kindle collection for download to her computer before Amazon’s restriction took place — she realized she wanted to know more about the metadata of her collection. She wanted easy ways to find groups of books, and to understand what she had.

Want more AI tips? Sign up for our Tech Today newsletter! 

The book list downloader extension

So she took to the Google and did a ton of searching. Her quest landed her on the Chrome Extension page of the Amazon/Kindle Book List Downloader, which, shockingly enough, is an Amazon Kindle book list downloader. 

I decided to try it, too. Why not? I’m always curious about what goes on inside the Gewirtz machine.

You can install it, and it will give you a short list of your Amazon assets as a demonstration. But if you want to unlock the full power of this nifty little tool, you’ll have to pony up a whole six virtual George Washingtons. Yeah, for $6 for unlimited uses, it’s a worthwhile investment. It’s always good to support innovative programmers.

Also: The best reading tablets: Expert tested and recommended

Installing the extension asks for the usual scary list of approvals. My advice is to disable this thing when you’re done using it, and only enable it for the few times you’re going to download a full list.

Next, click the puzzle-piece icon on your Chrome toolbar, then click the pin to pin the downloader to your toolbar.

You’ll see a little A icon, and when you click it, you’ll get a drop-down menu similar to this.

If you click the Get Books button, you’ll be given instructions about where you need to point your browser.

You can run this extension on two specific Amazon pages. If you run it on the first page shown, you’ll get only basic data, including ISBN/ASIN, URL, Type, Origin, Title, and Author(s). Origin appears to be where the book was acquired, but for me every record lists “purchase.”

Also: Kindle’s new AI recap feature helps bring you up to speed – how to try it

I recommend you choose the second URL, because that one generates a spreadsheet with a ton more interesting information. We’ll get to that in a few minutes, but first let’s do the downloading process and then move on to manipulating the data.

If you purchase the extension, you’ll want to put your serial number into the code field. I’m not entirely sure why the uses remaining shows “-4/null,” but it doesn’t seem to impact performance.

Make sure you’re on https://www.amazon.com/yourbooks. Then click the Get Books button. Please note that nothing will happen if you click again. Just scroll the drop-down a little bit and you’ll see this comforting little progress indicator.

You’ll know you’re done when you get a file-download dialog. Tell the downloader where you want your spreadsheet, and you’re done with this phase.

Looking at the spreadsheet

It turns out there’s a ton of information in this spreadsheet.

Data includes the following fields:

• ISBN/ASIN: This is the Amazon ID for the book.
• Link: Link to the sales page for the book.
• Acquired Date: When you bought it.
• RelationshipSubtype: Shows whether you purchased it, are sharing it, it’s part of the Kindle Owner’s Lending Library (KOLL), and possibly more.
• RelationshipType: For me, every item was listed as “owner.”
• Format: The format of the item; we’ll look more into this later.
• Title: The book’s title.
• Pages: The number of pages.
• Listening Length: For the Audible books.
• Genres: The item’s genre.
• Subgenres: This had no data for me.
• Series Position: If a book is part of a series, which book it is in that series.
• Series: Named book series.
• (First) Author: The main author’s name.
• All Authors: All authors listed for the item.

You can see how this might get interesting. You can do a lot with Excel and its data analysis tools.

I did think of using Microsoft’s Copilot within Excel to chop it up, but then I got a look at the price. I spend $99 per year for Microsoft 365 Family for all the apps and five users. But just to use the AI in Excel, I’d have to spend $20 per month — an extra $240 per year for a single added feature. That’s two and a half times the cost of the whole office suite. I’m sorry, but that ain’t right.

Also: How to use ChatGPT to summarize a book, article, or research paper

But then again, there’s ChatGPT. I do spend $20 per month for the Plus account, but ChatGPT provides a great deal of value. Not to mention my job involves writing about it every day.

So, let’s feed the spreadsheet to ChatGPT and see what magic it can produce.

Analyzing Amazon metadata with ChatGPT

Let’s start with a quick ChatGPT hint. ChatGPT gets very confused when the top row of a spreadsheet doesn’t contain the names of fields. You can cajole ChatGPT to pay attention to the right row by telling it the first value, but I’ve found things go much easier if you just chop off any unnecessary top lines and present ChatGPT with header names as the first row.

And that’s what I did. I then proceeded to upload the unmodified spreadsheet file instead of the modified file, thereby confusing it. As one does. I started a new session with ChatGPT and uploaded the correct file.

Also: Is ChatGPT Plus worth your $20? Here’s how it compares to Free and Pro plans

I started off playing around with the format field. I told it, “Tabulate the format field. Show me how many of each is in the file. Make me a pie chart.” I got back a nice pie chart, but it was a little messy.

I asked it to tidy it up a bit by prompting, “Use labels, but give me a legend for anything under 2%.” That turned out to result in a very workable chart.

As expected, the bulk of my reads are on Kindle. But I didn’t realize I had about as many paperbacks as I do audiobooks. Cool.

Also: 7 reasons Kindles are still a great buy, even without downloads

Next, we get to the area I was most curious about: genres. I instructed ChatGPT, “Tabulate the genres field. Show me how many of each is in the file. Draw a pie chart representing genre. Use labels, but give me a legend for anything under 2%.” The results were interesting.

It was not what I expected. First, I would have sworn that I had vastly more science fiction than either mystery or literature. But apparently, I bought a lot more mysteries. The young adult slice makes sense. There are a number of great book series, like Tom Swift and many books about steampunk and dirigibles that are aimed at younger readers. I do enjoy them.

Also: Want free ebooks? These 10 sites offer thousands of options

I have no idea how 3.4% could be romance. It’s possible some book that seemed interesting was in that category, my wife bought something on my account, or I downloaded it from a promotion for free in my bit-piggy phase. Who knows? Still, it’s informative.

There is more information I was able to tease out of that data with ChatGPT. It asked me, “Would you like this visualized or grouped further by theme or category?” My answer was yes, and I got back this.

I wanted to see if ChatGPT could give me a different take on my book list based on the titles themselves, so I asked, “Based on the titles of the books, show me a pie chart of the ten most popular themes.” I got this back.

Also: How to add EPUB, MOBI, and PDF files to your Kindle – no converting required

Weirdly, science or science fiction doesn’t show up on that at all, although science fiction is often lumped into the thriller or adventure categories. Time makes sense because I love time-travel stories.

And I’m again surprised by mystery’s prominence in this classification. If you combine mystery, mysteries, murder, and cozy, we get 56% of my reads are mystery-related. But I had no idea I had that many. I would have thought my Star Trek books would have smothered the other categories.

Also: How to easily convert EPUB files to Kindle format in minutes

Finally, I asked ChatGPT, “Looking at the spreadsheet, what other interesting information or insights can you surface for me?” This was fun.

It’s kind of interesting that my first Amazon book purchase dates back to 1998 — almost 30 years ago. It’s also interesting, since I read my Kindle books every day, that my last purchase was almost two years ago. Apparently, I’ve been working my way through my collection (which is, really, a good thing).

So there you go: a way to dig through all your Amazon reads and understand a little more about what you’re into.

Also: How to use ChatGPT’s Advanced Data Analysis to create quality charts and tables

What about you? Have you ever taken a deep dive into your Amazon book collection? Were you surprised by the number of genres or formats you’ve accumulated over the years? Do tools like this metadata exporter make you rethink your reading habits or spark new ideas for organizing your digital library? And if you’ve tried ChatGPT or similar tools to analyze your own data, how did it go? Let us know in the comments below.

Get the morning’s top stories in your inbox each day with our Tech Today newsletter.

---

You can follow my day-to-day project updates on social media. Be sure to subscribe to my weekly update newsletter, and follow me on Twitter/X at @DavidGewirtz, on Facebook at Facebook.com/DavidGewirtz, on Instagram at Instagram.com/DavidGewirtz, on Bluesky at @DavidGewirtz.com, and on YouTube at YouTube.com/DavidGewirtzTV.