A 10-K averages a couple of hundred pages, and most of it is boilerplate that hasn't changed since last year. The analysts who get through earnings season without drowning all use some version of the same trick: don't read the document — read the change. Here's the method, usable with any tooling (though we'll show where AnalystBook automates each step).
Step 1 — Start from what moved, not page one
Before opening the filing, know what changed quantitatively: revenue and margin moves, segment shifts, debt, cash generation. If you have the numbers pre-computed, this is a two-minute scan; the point is to walk into the text with questions. (In AnalystBook, the record and the what-changed report are already computed when the filing lands, and the Intelligence Brief's “What changed” section states the moves with figures.)
Step 2 — Diff the risk factors and MD&A
Companies edit Risk Factors and MD&A defensively — new paragraphs and changed hedging language are where trouble shows up first. Reading this year's section side-by-side against last year's is the single highest-yield 10-K habit. Doing it by hand means two PDFs and an afternoon; a materialized period-over-period diff shows you only the rewritten passages, ranked.
Step 3 — Go verbatim on the three sections that earn it
Skim nothing in: the accounting-policy and footnote changes (revenue recognition edits, new estimates, off-balance-sheet items), the debt note (maturity wall, covenants, new facilities), and anything the auditor touched (a changed auditor, going-concern language, critical audit matters). These are short, dense, and where the filing tells on itself.
Step 4 — Cross-check management's story against the signals
The narrative says “strong execution”; the computed signals say whether accruals are drifting from cash, whether adjusted EBITDA is drifting from GAAP, whether insiders are buying the story they're selling. Distress and manipulation scores exist exactly for this cross-examination — as flags that tell you where to dig, never as conclusions.
Step 5 — Write it down where you'll find it next quarter
The half-life of an unrecorded insight is about a week. Capture the read while it's fresh — what was good, what was bad, what you're watching — pinned to the passages that prompted it. Next quarter, the review takes half the time because your past self did the onboarding. This compounding is the whole reason we built the research workspace the way we did.
The 30-minute version, end to end
Two minutes on the computed changes → ten on the risk-factor and MD&A diff → ten verbatim on footnotes, debt and audit language → five cross-checking the signals → three writing the review. What used to be an afternoon becomes half an hour — not because you read less that matters, but because you stopped re-reading what didn't change.