Everything here had to be useful in the first eighteen months, which ruled out most of the famous business canon. Books about managing a division, defending market share or turning a good company into a great one are excellent and irrelevant when you are trying to find out whether anybody wants the thing at all.
The bias is towards books that tell you how to find out you are wrong cheaply. That is the actual work of the early stage, and it is unglamorous — talking to people, building the smallest testable version, and being honest when the answer is no.
Deliberately left off: anything whose central claim is that you need to raise money. Most businesses started in the UK do not, and a book that assumes a funding round quietly changes every decision downstream of it.