A supervisor once told me that a thesis fails twice: once when the argument breaks, and again when nobody can find where.

Lead with the claim

Academic writing rewards patience far less than people assume. State the claim, then earn it. A reader who reaches page forty still unsure what is being argued has already decided.

A structure that holds

1. The claim, in one sentence
2. Why the existing literature leaves room for it
3. The method, in enough detail to be repeated
4. What the evidence actually shows
5. What it does not show — this section is the one reviewers read closely

Write the limitations first

Counter-intuitive, but it works. Draft the limitations before the findings. It forces honesty into the argument early, and the findings come out narrower and far more defensible.

A claim you have already bounded is much harder to attack than one you defend after the fact.

On citations

Check that every cited paper says what you claim it says. Not the abstract — the paper. I have lost whole afternoons to a citation chain where the original study concluded the opposite of what three later papers attributed to it.