Before you write anything
Read the assignment brief twice, and the second time with a pen. Underline the task verb. Analyse, evaluate, compare, and discuss ask for different things, and a paper that describes when it was asked to evaluate will lose marks no matter how well it is written.
Then find the marking criteria. They are usually published and usually ignored. They tell you exactly what proportion of the mark goes to argument, to use of sources, to structure, and to presentation. That is a map of where to spend your time.
Step one: find the question inside the topic
You are usually given a topic, not a question. The first real task is to convert it. Ask what is genuinely disputed here, and what would count as evidence either way.
If the topic is the effect of standardised testing on teaching practice, the disputed question might be whether narrowing of curriculum is a consequence of testing itself or of how test results are used in accountability systems. That is arguable, and now you know what to read for.
Step two: read enough to have a position
Read before outlining, not after. An outline written before the reading is a list of guesses, and you will spend the drafting stage discovering that the sources do not support it.
Read abstracts and conclusions first, and go deeper only into the sources that survive. Take notes in your own words with citations attached. When you find yourself able to predict what a source will say before you read it, you have read enough to start.
Step three: outline in claims, not topics
Write your position in one sentence. Then write the two to four reasons that make it true. Those reasons are your body sections. If a reason needs more than one paragraph, it needs sub reasons, and those are your paragraphs.
This is faster than outlining by topic and it exposes weak arguments early. A reason you cannot state in a sentence is a reason you cannot defend in a paragraph.
Step four: draft the body first
Write the body before the introduction. The introduction announces an argument, and you cannot announce an argument you have not yet written. Most introduction problems are body problems that arrived early.
Draft badly and quickly. A first draft is a thinking device. Editing while drafting is why so many drafts stop after two paragraphs.
Step five: write the introduction and conclusion together
Once the body exists, the introduction writes itself: establish the question, indicate why it matters, state the position. The conclusion does the reverse: restate the position in light of what the body established, not in the same words as the introduction.
Write them in the same sitting so they agree with each other. A surprising number of drafts argue two different things at opposite ends of the same paper.
Step six: reverse outline
Print the draft. In the margin next to each paragraph, write one sentence saying what that paragraph argues. Read the margin sentences in order. That is your actual outline.
If a margin sentence repeats the one before it, merge the paragraphs. If a margin sentence does not connect to the thesis, cut the paragraph or rewrite the thesis. This takes twenty minutes and fixes more structural problems than a week of sentence editing.
Step seven: revise in passes
Fix one class of problem at a time. Structure first, then paragraphs, then sentences. Working on all three at once means rewriting sentences in paragraphs you will later delete.
Leave a gap before revising if you can. A day is enough to stop reading what you meant and start reading what you wrote.
A realistic time budget
For a two thousand word essay, expect roughly a third of your time on reading and planning, a third on drafting, and a third on revision. Most students allocate almost everything to drafting and then wonder why the finished essay reads like a first draft. It does, because it is one.