An account book is not a list of companies
Most target-account lists are exports. A few thousand rows, a column for industry, a column for headcount, and a hope that someone will work them. Nobody does. The list is too long to be a plan, so it becomes a place people search when they already have a name in mind.
An account book is the short version. For a narrow offer it might be every company that fits, and that can still be about a hundred. For a wide offer it is the ones you will actually work this quarter. The rest stay in a pool until something changes. A book you cannot finish reading is a list again.
What earns a place
A score is a sort order. A place in the book is a decision: this campaign, this proposition, and a person who owns the first conversation. You promote a company in. Search can suggest names that look like a fit. It does not promote them on its own.
Once a company is in, a few things belong on it.
The people you can already reach, and the roles where you still have nobody. "Three contacts, and none of them in procurement" is a useful sentence. A tidy tree with invented reporting lines is not. Public sources will give you the board and the named executives. They will not give you the reporting line under that, and guessing it produces a confident diagram you cannot defend.
The signals that would change the week. A tender. A contract that ends next year. A leadership change. An event you are both at. A signal goes in with its source and its date. If you cannot point at it, it does not go in the plan, because someone will eventually say it out loud to the client.
One live campaign. Two propositions at once is how the same person gets two different first messages from the same company.
A next step, with an owner. That part is the follow-up.
The pool can be large
The book should not be. Promotion is the moment a company becomes work, and work is what the daily 25 is for.
If you want a book sketched from a real website, the try page starts from a domain and reads only what that site publishes.