Happy Monday, everyone! Super short post today.

I’m very excited to invite all my readers to visit my scent library. This is my entire collection in all its glory. There are just over 250 fragrances on my shelf, which sounds like a lot until you realize that a sizeable portion of my menagerie are decants and minis. I’ve smelled almost everything on paper, but have worn only a a small fraction. That’s how this project came to be. I might acquire six, ten, or even more fragrances on any given purchase, and I needed something to manage and track them accurately. Roughly half of my collection ranges between 10ml and 30ml, with about 50 full size bottles (100ml). The rest are sample sets ranging from between 2-7.5mls. That being said, if I have the juice, and I can wear it or review it, it counts.

Why I Love Minis

I am not the type of collector/enthusiast who simply must own a full sized bottle. I value variety. Assuming a really good bottle of juice in the niche market sells for between $200 and $300 on average, I can usually find three or four decants on a discounter for that same dollar amount, really allowing me to stretch my budget. I am genuinely enthusiastic about my minis: receiving multiple decants in the mail instead of a single bottle is a genuine joy.

This is beneficial in any number of ways. First, they take up less space. Also, they won’t sit on the shelves for years; I am more likely to use the entire bottle. There is one drawback, though, of having so many perfumes: the paradox of choice. Sometimes, figuring out what I want to wear can be difficult. Every time I think I’ve decided, something else catches my eye and I think, “Oh, well, wouldn’t that be nice”.

I Had Help

Last thing, and full disclosure; while I gave Claude all the inputs, commands and dictated the structure, AI created this document. It also wrote the code that allows this document to be interactive, as the file is stored on another site. I had initially tried to create something similar on my own using Excel first, then Google Docs, and even MS word, attempting to utilize the table of contents features, which can be searchable. But it became unmanageable, time consuming and a little sloppy.

This was the perfect task for AI. Claude can read images, so this was literally as simple as taking a picture of my shelves, sample sets and decants, then uploading those to the engine. I told the machine what I was looking for, it showed me a sample, we tweaked it a couple times, and what you are looking at is the final product. This took a couple of days to get right, working up to two hours a day with my assistant. I am still concerned there might be some imperfections, and I really have to go through the document to check everything for accuracy, but I’m confident this is 90% there; maybe better, if I’m lucky. Please enjoy my collection.