Schools assign students to sections semi-randomly, often creating one nightmare class while others are fine. Teachers have no systematic way to predict or prevent toxic student combinations.
Ingests historical behavior data, grades, and teacher feedback to model student interaction effects. Recommends optimal section assignments that distribute high-impact students and avoid known bad pairings. Simulates outcomes before the semester starts.
Annual SaaS license per school ($2K-10K/year based on enrollment)
This is a genuine, visceral pain. The Reddit thread (4.4K upvotes) confirms teachers live this daily. One toxic combination can destroy an entire semester for 30 students and a teacher. The pain is seasonal (peaks during scheduling) but the consequences last all year. Counselors spend weeks on this manually. However, it scores 8 not 10 because the pain is accepted as 'just how it is' — many schools don't realize optimization is possible, so there's an education hurdle.
~130K K-12 schools in the US. Realistic addressable market is secondary schools (middle + high) where section assignments matter most: ~40K schools. At $5K average contract, that's a $200M TAM in the US alone. Realistic near-term SAM is more like $20-40M (tech-forward districts willing to adopt a niche tool). This is a solid niche business but not a venture-scale market unless you expand into broader scheduling or international.
This is the biggest risk. School procurement is brutal — long sales cycles (6-18 months), committee decisions, budget cycles aligned to fiscal years, and a bias toward bundled platforms over point solutions. $2-10K/year is reasonable per school, but getting the PO signed is the hard part. Counselors and teachers (who feel the pain) rarely control budgets. Administrators (who control budgets) may not feel this pain directly. You need to frame this as reducing disciplinary incidents, teacher attrition, and liability — not just 'better scheduling.' ESSER funds help in the near term but are expiring.
The optimization algorithm itself (constraint satisfaction + behavioral scoring) is buildable in 4-8 weeks for an MVP. The hard parts: (1) data ingestion is a nightmare — every school uses different SIS systems, behavior tracking tools, and data formats, requiring custom integrations or CSV imports, (2) behavioral interaction modeling requires meaningful historical data that many schools don't systematically collect, (3) FERPA compliance is non-negotiable and adds significant complexity to architecture, security, and contracts. A solo dev can build the optimizer, but the data pipeline and compliance layer are where the real engineering cost lives.
This is the strongest dimension. Nobody is doing this. Scheduling tools treat students as interchangeable. Analytics tools identify at-risk students but don't act on placement. ClassCreator is the closest but is a basic constraint solver without behavioral intelligence. The gap between 'we have behavioral data' (Panorama) and 'we optimize placement' (nobody) is wide open. The concept of modeling student INTERACTION effects rather than individual risk is genuinely novel in this market.
Natural annual renewal cycle perfectly aligned with school years. Schools reschedule every semester or year, so the tool provides recurring value by definition. Once a school's historical data is in the system, switching costs increase over time as the model improves. Behavioral data compounds — year 2 predictions are better than year 1, creating a retention flywheel. Annual SaaS contracts are standard in EdTech.
- +Massive, validated competition gap — nobody is applying behavioral intelligence to section placement
- +Pain is real, emotional, and well-documented (teachers and counselors viscerally understand the problem)
- +Natural recurring revenue with compounding data moats and increasing switching costs over time
- +Post-COVID behavioral crisis creates urgency and budget availability that didn't exist 5 years ago
- +Outcome simulation is a killer feature — letting administrators preview semester outcomes before committing is deeply compelling
- !EdTech sales cycles are 6-18 months with committee-driven procurement — cash flow will be brutal before product-market fit is proven
- !FERPA compliance and student data privacy concerns create legal, technical, and trust barriers that can kill deals even when the product works
- !Data quality dependency — the model is only as good as the behavioral data schools actually collect, which varies wildly and is often inconsistent or incomplete
- !Ethical and PR risk — 'AI labeling kids as disruptive' is one bad headline away from a school board banning the tool, regardless of actual methodology
- !Champion problem — teachers feel the pain but don't buy; administrators buy but don't feel the pain. Bridging this gap requires careful positioning.
Comprehensive SIS with master schedule builder that handles student-to-section placement based on course requests, teacher availability, and room constraints
SIS with scheduling module that assigns students to sections based on course requests, prerequisites, and capacity constraints
Student success platform that aggregates SEL surveys, behavior incidents, attendance, and academic data into dashboards for early intervention
Class placement software specifically for forming balanced classes based on academic levels, friendships, separations, gender, and special needs
Not a product but the actual incumbent: counselors and department heads using spreadsheets, Google Sheets, and institutional knowledge to hand-balance sections
CSV upload tool (skip SIS integration for V1). School uploads anonymized student data: prior behavior incidents, grades, teacher separation requests. System outputs optimized section assignments with a 'balance score' showing behavioral load distribution across sections plus flagged risky pairings. Include a before/after comparison showing current random assignment vs optimized assignment. Target 3-5 pilot schools with a free trial for spring 2027 scheduling. Keep it simple: constraint optimization with weighted behavioral scoring, not deep ML.
Free pilot with 3-5 schools (validate and collect testimonials) -> $2K/year per school for small schools, $5-10K for large ones -> Add district-level pricing at $3-5/student for multi-school deals -> Expand to include ongoing semester monitoring and mid-year transfer recommendations -> Eventually position as the behavioral intelligence layer that plugs into existing SIS platforms via API partnerships
6-10 months. Months 1-2: build MVP with CSV-based workflow. Months 3-4: pilot with 3-5 schools during their spring scheduling window (timing is critical — schools schedule for fall between March-June). Months 5-6: iterate based on pilot feedback. Months 7-10: convert pilots to paid and begin outbound sales for the next scheduling cycle. First meaningful revenue likely comes from Year 2 contracts signed during months 8-12. EdTech requires patience.
- “she also happened to be in the worst of my 6 sections I teach”
- “one guy transferred into my favorite class and then it became an absolutely horrific time for everyone”
- “This whole group was just pretty meh”