How to Write a Persuasive Essay: A Writing Instructor's Guide

I have read, at a rough guess, somewhere north of four thousand student persuasive essays. Most were competent. Very few persuaded me of anything, and the gap between those two facts is what this guide is about.

A competent persuasive essay has a thesis, body paragraphs with evidence, a counterargument, and a conclusion. It satisfies the rubric. A persuasive one changes, even slightly, what a careful reader believes when they put it down. That is a skill, and it can be taught, but not by handing you a template. What follows is closer to what I actually say in office hours when someone brings me a draft and asks why it got a B.

A Thesis Is a Bet, Not a Topic

The most common reason a persuasive essay fails is that it never risks anything. The student picks a position no sane reader would oppose, then spends five paragraphs proving it. “Bullying is harmful.” “Climate change is a serious problem.” These are not theses. They are weather reports. Nobody is going to write the rebuttal.

A real thesis is a claim a reasonable, informed person could reject. That is the whole test. If you cannot picture a thoughtful adult reading your first paragraph and thinking “I don’t think so,” you have written a summary of consensus, not an argument.

The fix is almost always to push the claim toward a decision. Not “social media affects mental health” but “schools should require a phone-free lunch period.” Not “the death penalty is controversial” but “states that keep the death penalty should require unanimous juries for every capital sentence.” The moment your thesis proposes that someone do something, or believe something specific about a contested question, it acquires opponents.

Opponents are what you want.

There is a second test, quieter than the first: can you support this with something other than your own conviction? Say a student argues that standardized testing should be abolished. Arguable, certainly. But when I ask what the evidence looks like and the answer is “everyone I know hates the SAT,” we have a problem. Belief is not evidence. Spend twenty minutes finding out whether the material exists before you commit. A slightly narrower thesis you can prove beats a bold one that collapses in paragraph three.

Write the thesis as one sentence and keep it in view. Every paragraph you draft afterward has to answer to it.

The Reader in the Middle

When I ask students who they are writing for, the honest answer is usually “you, so I can get the grade.” Understandable, but writing to an examiner means writing to nobody in particular.

Picture instead a specific reader who is undecided. Not the friend who already agrees with you, and not the ideologue who never will (unreachable in 1,200 words). The person you want is in the middle: reasonably informed, mildly skeptical, willing to move if you give them a real reason. Every choice in the essay, from what you explain to what you skip to which objection you cannot dodge, follows from who that reader is.

Suppose you are arguing for stronger tenant protections. If your imagined reader rents, you can assume they know what a sudden 30 percent rent increase feels like, and spend your words elsewhere. If your reader owns a duplex and rents out the other half, you had better acknowledge property taxes and repair costs early, or they will stop listening before your second paragraph. Same thesis, same evidence, completely different essay.

The most demanding version of this reader is the admissions officer, who reads perhaps forty essays in a sitting and has seen every angle on every prompt. If that is who you are writing for, our breakdown of the Princeton supplemental essays goes prompt by prompt through what each one is really asking.

Ethos, Logos, Pathos: Handles, Not Boxes

Aristotle sorted the ways a speaker moves an audience into appeals to character, to reason, and to emotion, and I have not found a better sorting since. What I have found is that students treat the triad as a checklist. Paragraph two gets a statistic (logos, tick), paragraph four gets a sad anecdote (pathos, tick), and somewhere in the introduction the writer mentions volunteering at a shelter (ethos, tick). The essay contains all three appeals and persuades nobody, because appeals are not ingredients. They are angles from which the same claim can be pressed.

Take one piece of evidence: a study showing that students who work more than twenty hours a week are twice as likely to drop out. That is logos, obviously. But how you introduce it is ethos: name the journal, note the sample size, mention one limitation, and the reader trusts both the study and you. What you place next to it is pathos: one sentence about what “twice as likely” means for a particular student on the edge. The same fact carries all three appeals. What changes is your handling. Logos still bears most of the weight in academic argument, but logos alone reads like a spreadsheet, and nobody changes their mind because of a spreadsheet.

Ethos, in a student essay, has little to do with credentials, since you probably do not have any yet. It comes from conduct on the page. Fair summaries of the other side. Sources named precisely. Concessions made without flinching. A tone that stays level when the topic is heated. I have watched strong arguments lose all their authority in one sneering sentence, because once a reader feels the writer is being unfair, every fact after that gets discounted.

Pathos is the appeal students both overuse and misuse. Overuse: the essay that opens with a paragraph of tragedy and never earns it with evidence. Misuse: reaching for outrage when something quieter would work. Recognition moves a skeptical reader. So does the slow realization that something is unfair in a way they had not noticed. Anger mostly does not, because it signals that the writer has stopped reasoning.

Structure That Gains Ground

I do not care much whether your essay has five paragraphs or nine. I care whether it builds. A persuasive essay should feel like walking up a slope: each paragraph a little higher than the last, so that by the conclusion the thesis looks less like a claim and more like the obvious place to have arrived. The conventional shape does this fine when it is used with intent.

Part	What it has to do	Where it usually goes wrong
Introduction	Establish the stakes, then land the thesis at the end	Opens with a dictionary definition or “Since the dawn of time”
Body reasons	One reason per paragraph, each with evidence and explanation	Evidence dropped in without a sentence saying why it matters
Counterargument	State the strongest objection fairly, then answer it	Picks a weak objection, or “answers” it by restating the thesis
Conclusion	Show what changes if the reader agrees	Repeats the introduction with the verbs swapped

The order of body paragraphs matters more than students expect. Lead with a strong reason so the reader takes you seriously, tuck the weakest in the middle where it does the least damage, and finish with your best material so the essay peaks right before the counterargument. That sequencing follows where a reader’s attention actually goes: high at the start, sagging in the middle, recoverable near the end if you give them a reason.

Inside each body paragraph the pattern is simple and almost nobody executes it fully. Claim, evidence, explanation. The explanation is the step that goes missing. A student quotes a statistic, moves on, and assumes the reader drew the connection. The reader did not. The sentence after the evidence, the one that says “this matters to my thesis because,” is where the persuading happens, and it is the one I most often write “so what?” beside. And if your paragraphs could be shuffled without loss, they are a list, not an argument: “even granting the cost, there is a second problem” carries the previous point forward where “another reason is” abandons it.

The Paragraph Nobody Wants to Write

Every semester somebody asks whether the counterargument is really necessary, since it seems to argue against the essay. It does not. It is the part that makes the rest believable, because a skeptical reader arrives with objections and waits for you to deal with them. If you never do, they assume either that you did not think of the objection, which makes you look uninformed, or that you hid it, which makes you look dishonest.

The paragraph works when you do two uncomfortable things. First, state the objection in its strongest form, the way its smartest advocate would put it. Choose a flimsy version so it is easier to knock down and an alert reader sees the dodge instantly. Second, answer it with something more than a restatement of your thesis. You can show the objection rests on a false assumption. You can concede part of it and show the rest of your argument survives. You can show the cost it raises is smaller than the cost of doing nothing. You can show it proves less than it seems to. What you cannot do is write “however, my position is still correct” and move on.

Concession is underrated. “It is true that a congestion charge falls hardest on low-income drivers who cannot easily switch to transit” costs you nothing and buys enormous credibility, provided the next sentence is “which is why the proposal exempts them and funds the bus routes they would use.”

One Thesis, Four Versions

Say a student wants to argue that community colleges should provide childcare. I will walk the claim from where it usually starts to where it needs to end.

Version one. “Childcare is important for student parents.” True and unopposed. The essay that follows will be five paragraphs of agreeing with itself.

Version two. “Community colleges should provide childcare for students.” Better, because now there is a should. But provide it how? Free? Subsidized? On campus? For which students? A reader cannot argue with it because they cannot tell what it commits to.

Version three. “Community colleges should offer free on-campus childcare to any enrolled student with a child under five.” Now it is arguable. A trustee could say the college cannot afford it; a taxpayer could ask why parents get a benefit others do not. Those objections are the essay’s raw material, and the student can now look for evidence: dropout rates for student parents against the general population, completion data from campuses that already run childcare centers, the tuition revenue lost when a parent withdraws in the first semester, and what comparable programs actually cost.

Version four is the same thesis with the argument attached, and the difference between three and four is the difference between an outline and an essay. The introduction opens on a specific hour, say 11 p.m., when a parent finally sits down to a statistics assignment after a child is asleep, and it lands on the thesis in its final sentence. The first body paragraph establishes the scale of the problem with enrollment and completion figures, and then, in the sentence students always forget, says what those figures mean: a large share of the college’s own students are at higher risk of leaving for a reason the college could address. The second body paragraph shows that campuses with childcare see measurably better completion, and it takes the trouble to explain why that is more than correlation, which means the student has to find a study that controls for income and prior achievement and has to say so. The third addresses cost head on, and this is where the counterargument lives: the trustee’s objection about budgets, stated as the trustee would state it, answered with the revenue lost to withdrawal, which on many campuses exceeds what a childcare center costs to run. If the numbers do not support that answer for the student’s own college, the student says so and argues for a pilot instead, a smaller claim but an honest one. The conclusion does not repeat the thesis. It asks what the college is for, if not to make finishing possible for the people who most need to finish, and leaves that question with the reader.

Revision, or Where the Grade Actually Lives

What I tell students about first drafts is this: they exist to get the argument out of your head and onto the page where you can see it. Grades come from what happens next.

Read the draft as a person who disagrees with it. Not a hostile reader, a fair one who would like to be convinced and has not been yet. Ask, at every paragraph, whether it moved you. If a paragraph exists because the outline said there should be one there, cut it or rebuild it. If a claim is louder than its evidence, find the evidence or turn the volume down. If the counterargument made you nod along with the objection and then feel let down by the answer, that is where your remaining hour goes.

Then read it aloud, literally. Your ear catches the sentence that runs out of breath, the transition that does not transition, the paragraph that opens the same way as the one before it. Your eye skims all of that. Fix the logic before the prose; there is no point polishing a sentence you are about to delete.

A few things I mark in the margin so often that I keep the phrases ready.

• Hedging. “It could perhaps be argued that” tells me you are not sure. Qualifying a claim precisely is good; apologizing for it is not.
• “Studies show.” Which studies? A source you cannot name is not a source.
• Absolute words. “Obviously,” “clearly,” “everyone knows.” Each says to the skeptical reader: I am not going to bother convincing you.
• A conclusion that restates. If your last paragraph could be swapped with your first and nobody would notice, you have not concluded anything.

Persuasion on the page is a slow thing. You are trying to be the kind of writer a careful reader is willing to follow for twelve hundred words, and to have somewhere worth taking them. Pick a claim you can be wrong about. Find out whether you are. Then make the case.