Admissions · What changed

The quiet shift in what "impressive" now means on a STEM application

Strong grades and test scores used to set a capable engineering-minded kid apart. A decade of doubling applicant pools changed that. Here's what admissions readers say they're looking for instead — and one honest, hands-on way to give a smart kid something real to show.

If you're the parent of a bright, STEM-leaning high schooler, you already sense it, even if no one has said it plainly: the transcript that would have stood out ten years ago now barely clears the bar. It isn't your imagination, and it isn't your kid. The math simply changed underneath everyone.

Consider one field. The number of computer-science bachelor's degrees conferred in the U.S. more than doubled in a single decade — from roughly 51,700 to about 112,700 between the 2013–14 and 2022–23 academic years. When a pool doubles, the credentials that used to signal "exceptional" quietly become the price of entry. A 4.0 and a strong test score don't distinguish a student anymore. They qualify them to be compared.

What admissions readers started rewarding instead

As grades compressed toward the top, selective-admissions counselors began describing the student they actually want with a word that sounds almost like an insult: not well-rounded, but "well-lopsided." Angular. Pointed. A kid who went unusually deep on one real thing, rather than collecting a shallow layer of everything.

The people who read applications say a version of this out loud. MIT's admissions team reminds applicants that there are many ways to make an impact — you don't need to have cured a disease by fifteen for your work to count; what matters is that it's genuine and it's yours. A former Vanderbilt admissions officer has written that the students easiest to champion in the room are the ones with a genuinely interesting project, because a real project hands the reader a story they're excited to tell.

Notice what that is, and what it isn't. It isn't a guarantee. It's evidence — the difference between telling someone your kid is capable and showing them a working device your kid built, wired, debugged, and can explain. That distinction is the whole game now.

The trap most well-meaning families walk straight into

So a parent goes looking for "a project." The internet points them at a giant electronics kit — two hundred components, thirty-plus mini-projects, a big satisfying box. It photographs beautifully. It also, very often, ends like this:

Walk any maker forum and you'll find the same quiet confession from families who bought the big kit: we did the tutorials, it was fun, and then… now what? They followed the instructions, they blinked the LEDs, and at the end they had no single original thing to point to. The breadth was the problem. Thirty shallow starts don't add up to one finishable project — and an unfinished project is worse than none, because it teaches a kid that they don't finish.

Depth over breadth — one build, done all the way

The opposite approach is almost boring in its simplicity: one focused, genuinely advanced build, taken all the way to done. Not a sampler. A single real device — say, an RFID access-control system: the kind of card-reader lock you tap a badge against — with every part needed for that one build and nothing left over to lose interest in.

Depth is what's legible to an admissions reader, and depth is what survives the interview when someone asks a follow-up question. "I built an access-control system, here's how the reader talks to the microcontroller, here's the bug that cost me an evening" is a conversation. "I have a kit" is not.

The part almost nobody includes

Here's the gap even the good kits leave open. A finished device, sitting on a desk, is not yet a portfolio project. It becomes one when it's written up — a clean project summary, a couple of real photos, a short honest account of what was built and what was learned. That documentation layer is the thing that actually travels into an application, an activities list, or an interview.

This is where the whole thing comes back to a number. Families in this position already treat enrichment as a normal line item. The average comprehensive college-consulting package runs about $6,500, and independent consultants now commonly charge $140 to $600 an hour. Against that backdrop, a one-time kit that produces a real, finished, documented project isn't an expense to agonize over. It's a rounding error that happens to leave behind an artifact.

Sources referenced: National Student Clearinghouse Research Center (2024); IECA / industry consulting-fee reporting; published admissions guidance from MIT Admissions and college-counseling professionals. Statistics are cited as evidence of a market shift, not as a claim about any individual outcome.