Why I Started Tracking Every Book I Read

A few years ago I started writing down every book I finish — title, date, a few lines on what stuck. It sounds annoying or completely dorky. It turned out to be extremely high leverage, helping me to remember the books and knowledge gained as well as overarching themes.

What the log revealed

I read that even while president George W Bush and Rumsfeld had annual reading competitions often reading more than 50 books a year despite a hell of a lot of work. That inspired me to read more and as I began reading a book a week I started keeping track of what I read so that I didn't repeat. That's embarrassing to admit but sure I was now "reading" a lot but I couldn't remember them well enough to hear a books title and be sure if I had actually read it or not. A list helped with that.

The second surprise was retention. The act of writing a sentence about a book — not a summary, just what stuck — does something a highlight never does. It forces a verdict. A year later, I remember the books I logged and have functionally lost the ones I didn't, including some I'm sure I enjoyed more. Now I often run the lists through an LLM to create more detailed summaries and read them as well.

The third surprise was the patterns. Scan five years of entries and you see your own mind from the outside: the year everything was about energy markets, the slow drift from strategy books toward biographies, the fiction drought that perfectly tracked a brutal deal cycle. The log is a mirror with a long memory. The AI summaries and insights provide even more data and ability to analyze changes over time.

The system, such as it is

People ask about the system expecting software. Here it is in full:

1. One line per book: title, date finished.
2. One sentence on what stuck. If nothing stuck, write that — it's the most useful entry of all.
3. One flag: would I reread this? The "yes" list, after a few years, is the real library.
4. If I quit the book (which is rare) write that as well and why.

That's it. No ratings, no tags, no app. The friction has to be near zero or the habit dies by February.

The meta-lesson is the same one that shows up everywhere from portfolios to org charts: the cheapest way to get better at something is to start keeping score. The score doesn't even need to be good. It just needs to be written down.