Last reviewed
Correct answer: D. Each smaller prompt is easier to execute well and easier to check along the way
Explanation
A long task fails as one thing. You get a finished output that is wrong somewhere in the middle, and the only repair on offer is to run the whole prompt again and hope. Splitting it — outline, then draft one section, then check that draft against the brief — changes both halves of that. Each prompt carries less at once, so it is executed better; and each output is small enough to actually read, so a mistake is caught at the step that made it.
It pays most where errors compound. If the first step selected the wrong ten rows, everything after it is careful work on the wrong data, and no amount of quality in the later steps recovers that.
What it does not buy is worth stating. Chaining usually costs more total tokens and more elapsed time, not less, because the steps run in sequence and each one repeats some context. And it does not remove review — it relocates it, from one hard read at the end to several easy ones along the way.
Sources
“Explicit prompt chaining (breaking a task into sequential API calls) is still useful when you need to inspect intermediate outputs or enforce a specific pipeline structure.”
Practise 4 questions on this topic
Take Advanced Prompting Techniques — Timed Test (4 questions) — scored instantly, explanation for every question, no login.