There is no shortage of ACT content. There are thousands of practice questions, hundreds of video explanations, entire books dedicated to every section. The content problem has been solved for decades. So why does ACT prep still feel like it was designed to be endured rather than used?
Because it's a product problem. And nobody in the space is treating it like one.
The diagnosis problem
Most platforms throw content at you and call it personalization. You take a practice test, get a score, and then get handed a list of topics to "review." That's not diagnosis. That's just sorting. Real diagnosis means understanding why you're getting a question wrong — not just which category it lives in.
Are you missing Reading Comp questions because you're misreading the question, running out of time, or making careless inference errors? Those are three completely different problems with three completely different fixes. Every existing platform treats them the same. They all say "work on Reading Comprehension." That's useless.
A score isn't feedback. It's just a number. Feedback tells you what to do next.
The second issue is that practice and diagnosis are decoupled. You drill problems, then separately you get told what you got wrong. There's no feedback loop inside the actual study session. You have to wait until after to know if you understood anything — and by then you've moved on mentally.
What I'm building instead
36.fyi starts from the assumption that the problem isn't content — it's delivery. Same underlying questions, but the product wraps them in a diagnostic layer that actually tells you something actionable.
Phase 1 (live now): the core loop. Practice questions, real-time feedback, basic tracking. It's not glamorous but it's the foundation. Everything in Phase 2 is built on top of the data that Phase 1 generates.
Phase 2: proper weak-spot identification. Not "you struggle with math." Specific: you lose points on multi-step algebra problems specifically when there's a variable in the denominator. Then targeted drills for exactly that. Not 40 questions — maybe 8, chosen precisely.
The scholarship angle is real. A 34 → 36 jump can mean $40,000+ in merit aid at certain schools. That's not a tutoring business — that's an ROI conversation. The product has to earn that framing.
Phase 3: the scholarship unlock layer. Because a higher ACT score isn't just a number — it's money. And high schoolers who understand that study differently. I'm building tools to show you exactly what score you need for what scholarship at which school, so the practice has stakes beyond "do better."
Why nobody has done this
Honest answer: the existing players are legacy businesses. They built their platforms in a different era, they have customer bases that make switching costs real, and they're not incentivized to rebuild from scratch. They iterate on the margins.
The other reason: you have to actually understand the ACT to build good tooling for it. Not just "I know the test exists." I mean you have to understand the specific error patterns, what the College Board actually tests, why certain wrong answers are designed the way they are. Most edtech founders don't have that context. They outsource content creation to tutors and build a generic wrapper around it.
I'm inside the problem. I'm the customer. That's not a credential — it's an information advantage.
I'm inside the problem. I took the test. I know what the bad study sessions feel like. I know what it's like to open Khan Academy and immediately close it because nothing there maps to what I actually need. That's not a credential — it's an information advantage, and I'm using it.
The $1B+ test prep industry is full of content. It's starving for product thinking. That's the gap I'm in.
If you're a high schooler trying to move your score, 36.fyi is live. Phase 1 is the start. More coming.