What we believe about building edtech products that actually work. These principles shape every skill in this repo and every company that goes through ScaleU.
Informed by "Cracking Higher Ed: Why Startups Miss the Mark" (SXSW EDU 2026, CC BY 4.0).
"Interesting" and "funded" are completely different categories. A waitlist is not demand. A dean saying "transformative" is not demand. Demand is a budget holder who can name the line item, a procurement process that started before the pilot ended, and a product that survives if your champion leaves tomorrow.
Before you build, answer five questions: Who is struggling, and what is the struggling moment? What have they already tried? Is there a budget line item? What happens if they do nothing? Who else must say yes? If you can't answer all five with specifics, you have a hypothesis, not a market.
Pain points show up in one phase of the student journey but originate 1-2 phases earlier. Students drop courses in week 2, and founders build retention dashboards. But the real failure point was months earlier, when nobody translated "7.5-week accelerated session" into "15+ hours per week on top of your job."
When you hear a problem, ask: what happened one phase earlier that made this inevitable? The upstream job is usually less crowded and higher-leverage. Build for the cause, not the symptom.
Most edtech founders overestimate their evidence tier. A teacher saying "my students loved it" is not evidence. Usage data showing high engagement is not evidence of learning outcomes. Positive pre/post scores without a comparison group are not Tier 3.
Know where you stand on the ESSA evidence ladder (Tier 1-4). Know what tier your buyer requires. Have a concrete plan to close the gap. Investors and institutional buyers increasingly demand real outcome data, not testimonials. The skills in this repo cite 376 peer-reviewed papers because evidence is the language that institutions trust.
Students describe "teaching myself" and "absent instructors." Faculty independently describe "unsustainable class sizes" and "burnout." Institutions track these as separate problems in separate reports. They are the same structural constraint described from two positions.
The tool that solves the provider's capacity job often solves the student's experience job as a byproduct. And the provider side is where the budget and procurement authority sit. When you hear a student experience problem, find the mirror on the provider side. That's where the purchase order comes from.
Every stakeholder tells you the same thing: "X is the problem." Then the outcome data says X actually produces better results. This happens more often in education than you'd expect. Qualitative data overrepresents the people who struggled. The silent majority who succeeded leave no trace in interview data.
If every stakeholder agrees on the problem, that's a signal to validate quantitatively, not confirmation to build. Unanimous qualitative agreement can be a bias artifact. Check the outcome data. If it contradicts the narrative, the real opportunity may be different from what everyone is asking for.
"We sell to universities" is not a market. Name a specific person at a specific institution with a specific problem. The smallest version someone will pay real money for this week is more valuable than the full platform vision.
Your first 10 customers should look alike. Same institution type, same buyer persona, same problem. Prove it works there. Then expand from strength. Trying to be everything to everyone in education means you're nothing to anyone.
80% of edtech founders build for Course Experience, the most crowded phase of the student journey with 15+ product categories competing. The biggest opportunities are in the phases nobody is building for: Pre-enroll, Apply, Onboard, Select & Enroll.
If everyone is building in the same space, that's not validation. That's a warning. The pain in those underserved phases is quantified at scale but product coverage is thin. Look where others aren't looking.