Three decades ago, if you wanted a trade book, you went to a bookstore or a bigger merchant like Wal-mart or a department store with a book “section”. It was actually hard to get a book any other way. That really changed starting with Amazon in 1995 and has continued to splinter since with a large number of online retailers making books available that they (mostly) source through Ingram.

So we have gone from a world where about 95 percent of books were bought in stores to a world where that number is now more like 20 percent (by my guestimate.)

When you think about it a little bit, that makes all the sense in the world. Books must be the very best single consumer product to purchase online rather than in a physical shop. For five big reasons.

1. The choice of book titles has become so vast — going from about 500,000 titles in English three decades ago to about 20 million titles available through Ingram today — that the chances of finding any particular book you are looking for in any particular store is vanishingly small.

2. You don’t need to try on a book or look at it closely to know what it is. In most cases, the cover that you see, the description you can read, including its metadata, will tell you what you’re getting so that the chances of surprise or disappointment is also vanishingly small.

3. You almost never need any particular book “right now”. Almost always, 2 or 3 days from now is just fine. (And you wouldn’t find it “right now” in a store most of the time anyway.)

4. Books are heavy. You don’t want to carry one (let alone two or three) around with you if you aren’t coming straight home from the store. Getting it delivered to your home is a much more convenient option.

5. Price variation is limited. Books usually have prices set by the publisher which determines pretty much what you’re going to pay for it, regardless of where you buy it. (And online can be cheaper some of the time, anyway.)

So, if you know what you want, the chances of finding it in a store are too small to be worth the attempt. And if you don’t know what you want, the range of choices you’ll find online dwarfs what you’ll find in any stores.

It is true that some people like being in a room full of books. It is true that some people like having the choices of possible reads “curated” for them. And it is also true that the book might be purchased as a gift, so you do (in that case) need something right now. There are people shopping for illustrated books and children’s books where there can be value added by seeing the book up close. But those customers constitute the bookstore audience today. Thirty years ago every purchaser of a book was the bookstore’s audience.

It is not Borders’ fault that they disappeared. And it isn’t Barnes & Noble’s fault that their shelf space (if not their number of locations) continues to shrink. Quite aside from the sales lost to ebooks and audiobooks (and they aren’t insubstantial), bookstores have a nearly insuperable challenge creating a compelling reason to use them.

There will always be bookstores, but they’ll be increasingly rare and idiosyncratic.

My friend Thad McIlroy has written a brilliant piece for PW about the impact of AI like ChatGPT on book publishing. In it, he suggests that it could kill trade publishing. I loved Thad’s piece, but, to my way of thinking, “trade publishing” is already just about dead. After all, the term was invented because the books were sold through the book trade, and I have long believed that the book trade is no longer capable of supporting much publishing. Thad’s piece motivated me to have a chat with him, which resulted in me delivering him this summary about the bookstores. He encouraged me to put it writing. I encourage anybody reading it to read his piece on AI.