The essay is the one place a STEM file argues in prose instead of numbers. We coach it the way we coach a derivation — find the claim first, then build the support. This guide is our method, shown in full.
Every school phrases its supplements differently, but strip away the wording and almost all of them are variations on a small set of questions. Learn the set, and a wall of prompts becomes a handful of arguments you already know how to build. For each archetype we give the same three things: what it is really asking, the way it usually fails, and the angle a STEM applicant is best placed to take.
Have you done the reading, and can you name specifics only this program offers — a lab, a course sequence, a way it teaches — that map onto something you already do?
Praise that would survive a find-and-replace of the school's name: "world-class faculty," "vibrant community," "endless opportunity." Interchangeable flattery reads as no research at all.
Name a professor's actual line of work, a specific course, or a lab, and connect it to a project you have already started. Fit is a claim you prove with a concrete plan, not a compliment.
What do you give to a group of people, not just take from it? Admissions reads this as a forecast of how you will act as a roommate, a lab partner, a classmate.
Defining community so broadly ("humanity," "my generation") that no action fits inside it — or listing memberships without a single moment where you changed something for someone.
The math team you organized, the robotics kit you taught younger students to solder, the wiki you kept for a club. Technical people contribute by building the thing everyone else uses; show that.
What do you chase when no one assigns it and no grade is attached? The reader wants evidence of a mind that keeps going after the bell.
Declaring passion in the abstract — "I have always loved science" — without a single question you actually pursued, or narrating a class syllabus back to the school that wrote it.
Track one real question from the moment it caught you to what you did about it: the paper you struggled through, the simulation you coded, the follow-up the answer opened. Curiosity is a trail, not a trait.
Behind the one-line entry on your activities list, what did the work actually feel like, and what did it make of you? The prompt wants depth the list cannot hold.
Re-listing responsibilities the reader already saw, or choosing the most impressive-sounding activity over the one you can write about with real texture.
Pick the project you can render in scenes — the bug that cost a week, the experiment that failed twice before it worked. Specificity about process is what separates a builder from a résumé.
What context shaped how you see, and what will you therefore notice that others in the room will miss? This is about the lens, not the label.
Naming a background and stopping there, as if the category speaks for itself — or reaching for hardship you did not live because the prompt seems to want it.
Let the background do work: the bilingual habit that makes you good at translating between a client and an algorithm, the household constraint that taught you to build with what was on hand. Perspective is shown by what it lets you do.
Can you look honestly at something that went wrong and show a mind that revises? The growth is the point; the setback is only its setup.
The humble-brag failure ("I worked too hard"), or a catastrophe narrated for sympathy with no reflection — all storm, no learning. Readers discount both on sight.
Debugging is a failure story you already know how to tell: a wrong hypothesis, the evidence that overturned it, the corrected model. Treat a personal setback the same way — claim, disproof, revision.
Favorites, lists, one-liners — a fast read on personality and judgment. In a handful of words, is there a real person with real taste?
Straining to be funny, or playing it so safe the answer is invisible. Both waste the one prompt designed to be effortless and human.
Answer honestly and let the specifics carry the charm — the odd tool you love, the problem you re-solve for fun. Precision is its own kind of voice; you do not have to perform whimsy.
Good essays are not written so much as argued into shape. We run the same five-step process on every file, in order, and we do not skip ahead — the discipline is what keeps a draft from becoming a pile of nice sentences with no spine.
Before a word is written, we read everything: the student's transcript, activities, and drafts on one side; each school's official prompts, word limits, and prompt counts on the other. You cannot argue for a case you have not read, and you cannot answer a question you have not looked at closely.
Every essay must land exactly one claim — the single thing a reader should believe about the student when they finish. If we cannot state it in a sentence, the essay is not ready to be drafted. Everything that follows either serves that claim or gets cut.
A claim is only as strong as what supports it. We list the concrete scenes and specifics that prove it — moments, objects, exact decisions — not adjectives. "Determined" is an assertion; the night you re-ran the assay a fifth time is proof.
The first draft is written long and unguarded, in the student's own voice, without worrying about the limit. Then the real work begins: subtraction. Most essays get better as they get shorter, and the discipline of the word count is a feature, not an obstacle.
A finished draft has to earn its place in the whole application. We ask the plain question: does this essay do the job the file needs — say something the transcript and activities cannot, and answer the exact question the school asked? If not, we return to step two.
The most common way strong STEM applicants sink their own essays is by writing them for the wrong reader. An admissions officer is intelligent and generous but rarely a specialist in your subfield, and an essay dense with acronyms and unglossed methods reads as a wall, not a window. The goal is not to dumb the work down; it is to make it legible — to let a smart non-specialist feel the stakes of the problem and see your hand in solving it.
We coach three moves. First, lead with the question, not the apparatus: say what you were trying to find out and why it was worth finding out, before any technique appears. Second, translate every method into its purpose — a reader does not need the name of the algorithm, they need to know it let you separate signal from noise. Third, keep yourself in the frame: the essay is not a methods section, so the moments that belong are the ones where you decided, doubted, or changed your mind. That is where intellectual vitality actually shows — in the thinking, not the terminology.
The strongest STEM files read as one story, not a scatter of unrelated activities. We call this the evidence chain — a traceable arc from foundations to customized research, where each piece explains the next: the class that raised a question, the competition that sharpened it, the project that finally chased it down. When the essays trace that chain, a reader stops seeing a list of accomplishments and starts seeing a mind that has been heading somewhere all along.
Planning that starts from last year's questions is planning built on a guess. We maintain an internal database of the official supplemental prompts for more than 100 institutions — national universities and liberal-arts colleges alike — capturing each school's exact wording, word limits, and number of required and optional essays. We refresh it every cycle, because schools add, cut, and rewrite their prompts more often than families expect.
Because we work from this year's actual questions, a student's essay calendar is real from day one: how many pieces each school demands, which prompts overlap enough to share a spine, and where a single strong argument can be adapted across several applications. It is our own reference of public prompts — not a corpus of anyone else's writing — and it exists so the strategy is grounded in fact rather than folklore.
An essay is only worth anything if it is true and it is the student's. Everything we refuse to do follows from that one commitment — the work has to be theirs, argued well, or it is not worth putting a name to.
Readers see thousands of essays a season and recognize the manufactured ones instantly: the borrowed structure, the too-even cadence, the sentences no seventeen-year-old would actually write. A file that reads as assembled gets discounted before its content is even weighed.
Find the claim first. Then build the support. The essay is a derivation, not a decoration.
— The Tian2 EditorsA free editorial showcase. We share method openly; there is no storefront here.
This guide is our method as we actually run it, worked from official prompts and written in our own words. ← Back to STEM College Counseling