Source: competitive_landscape.md
Research date: March 2026
What it does: An AI-powered college application assessment platform that uploads a student's Common App PDF and transcript, extracts academic and extracurricular data automatically, and returns a holistic score on a 0–240 scale with a national percentile rank and school-specific competitiveness estimates.
Target user: High school students targeting selective and Ivy League institutions.
Key features: - Automatic extraction from uploaded Common App PDF - Five-factor scoring: Academic (GPA/rigor), Personal (EC impact), Alignment (coherence of course + activities narrative), Awards, and Testing (SAT/ACT/AP/IB) - Score band interpretations: 220–240 = HYPSM-competitive; 200–219 = highly selective; 180–199 = competitive; below 160 = limited selective options - National leaderboard for self-comparison - Personalized improvement feedback
Pricing: Completely free (no subscription, no payment tiers mentioned).
Data sources: "Based on informal conversations with Ivy League admissions representatives and published institutional research." Explicitly uses "Ivy League methodology established in 1985." The platform itself disclaims: "should not be considered official admissions guidance."
Limitations: - Score is a holistic estimate, not a per-college probability - Relies entirely on user-provided document accuracy - Does not model college-specific admission functions or competitive applicant pools - No round-specific modeling (ED/EA/RD) - No yield or enrollment decision modeling - Score bands are broad and do not differentiate between, e.g., Harvard vs. Cornell
Differentiation from ABM: academicindex.ai scores a single student in isolation. It cannot answer "given the applicant pool this cycle, what fraction of students with your profile will actually enroll at Harvard?" because it has no model of the competing population.
What it does: A freemium platform with a probabilistic chancing engine that estimates per-college admission probability using a student's academic profile, extracurriculars, demographics, and school context. The engine is calibrated against actual admission outcomes from CollegeVine's user base.
Target user: High school students (self-service); also sells peer essay reviews and paid advising.
Key features: - 75-factor analysis including GPA, test scores, course rigor, EC tier (12-sub-tier taxonomy from Tier 1a rarest to Tier 4c most common), gender, geographic location, intended major - Calibrated accuracy: when the engine predicts X%, approximately X% of users with that score actually got in (48.1% admitted when predicted at 50%) - What-if simulator to see how profile changes affect chances - College list builder with safety/target/reach categorization - Peer essay review marketplace
Pricing: Freemium. Core chancing engine, college search, and community Q&A are free. Essay reviews ~$60 each. Full advising packages low thousands.
Data sources: Self-reported outcomes from CollegeVine's own user base (millions of students). Supplemented with Common Data Set figures. Not institutional or official.
Limitations: - Does not model essays, recommendation letters, or high school context (explicitly acknowledged) - Does not model ED/EA round multipliers in the current iteration - Does not model legacy, donor, or athlete hooks - Pool is self-selected CollegeVine users — likely over-represents engaged, higher-resource applicants - Cannot model how results at one college affect yield and waitlist movement at others - Individual predictions are probabilistic estimates, not deterministic
Differentiation from ABM: CollegeVine computes P(admit | student profile). An ABM computes P(enroll | student profile, entire competing pool, sequential round dynamics, yield cascades). CollegeVine is a slice of the decision tree; an ABM is the whole tree.
What it does: Premium admissions consulting firm ($15,000–$30,000+ per student) with a free lightweight admissions calculator as a lead generation tool. The calculator takes SAT/ACT and GPA inputs and returns a bucketed list of Safety, Target, and Reach schools.
Target user: High-resource families willing to pay for one-on-one expert coaching. Crimson Rise targets middle schoolers (grades 6–8) for early pipeline development.
Key features: - Free calculator: inputs SAT/ACT + GPA; outputs school tier buckets (Safety/Target/Reach) - AI Session Summary tool: automates note-taking after consulting sessions, sends summaries to students and parents - Claimed 98% acceptance rate to students' top choices; 35% average acceptance rate at top-15 U.S. institutions vs. ~5.6% general population - Access to former admissions officers as consultants
Pricing: Free calculator (lead gen). Full consulting services: $15,000–$30,000+ for undergraduate admissions support.
Data sources: Calculator relies on basic GPA/test score benchmarks. Consulting draws on institutional knowledge of former admissions officers.
Limitations: - Free calculator is rudimentary (GPA + test score only, no EC, no hooks, no round) - Heavy paywall for any meaningful guidance - Success rates are marketing claims, not independently audited - Accessible only to very high-income families - No systemic or population-level modeling
Differentiation from ABM: Crimson is a human services business using lightweight tools as acquisition funnels. An ABM produces quantitative, replicable simulations of system-level behavior that any student or counselor can explore at zero marginal cost.
What it does: A micro-scholarship platform where students accumulate scholarship credits from partner colleges by logging ongoing achievements (grades, clubs, sports, volunteer work) in grades 9–12. Not primarily a chancing tool.
Target user: Students seeking financial aid, particularly from less-selective partner colleges; colleges seeking enrollment marketing reach.
Key features: - Achievement portfolio building from 9th grade onward - Micro-scholarships auto-calculated per achievement based on partner college rules - 300+ partner college network - Average student earns ~$25,000 in micro-scholarship credits over four years - Free for students; colleges pay for enrollment marketing access
Pricing: Free for students. Revenue from partner college subscriptions.
Data sources: College-defined scholarship award rules. No admissions probability modeling.
Limitations: - Not an admissions predictor — only models financial incentives from partner colleges - Partner colleges are predominantly less-selective institutions (not HYPSM/Ivy+) - No chancing, no simulation, no competitive pool modeling
Differentiation from ABM: RaiseMe operates in financial aid / enrollment marketing, not admissions prediction or market simulation. Minimal overlap.
What it does: A school-licensed college and career readiness platform used by high school counselors to manage the college application process for their students. Includes scattergrams (GPA vs. SAT plotted against past outcomes from that school to specific colleges), college search, and application tracking.
Target user: School counselors (primary); students and parents (secondary). Sold as a school/district license.
Key features: - Scattergrams: plots historical GPA/SAT outcomes from your specific high school to each college (admitted/denied/waitlisted) - AI-powered "PowerBuddy" for personalized college and career discovery - AI recommendation letter drafting tool for counselors - Application tracking and document delivery - Career interest inventory - Common App integration for transcript and document submission - 2025–26 updates: work-based learning matching, real-time admissions data feeds, redesigned counselor dashboard
Pricing: School/district license (per-student pricing, not publicly disclosed). Not available to individual students. Typically in the range of $10–$30 per student per year based on industry norms.
Data sources: Historical outcomes from your own high school's Naviance data. School-specific, not national. Supplemented with Common Data Set and college self-reported information. PowerSchool acquired EAB's Naviance in 2019.
Limitations: - Scattergrams are school-specific — a school with 5 applicants to MIT has meaningless data - No national applicant pool visibility - Does not model hooks, legacy, donor status, or round-level probabilities - Cannot model enrollment decisions, yield, or waitlist dynamics - Data accuracy depends on counselor data entry quality - Individual school scattergrams are not calibrated probability estimates
Differentiation from ABM: Naviance shows you where your school's past students landed. An ABM shows you how the entire national applicant pool sorts itself — and what happens to your specific cohort under different admission policy scenarios.
What it does: A college and career network connecting students, high school counselors, and colleges through an application platform with an AI chancing engine ("Admission Intelligence") trained on tens of millions of de-identified historical outcome records.
Target user: High school counselors (primary); students; colleges for enrollment marketing.
Key features: - Admission Intelligence: ML-based chancing powered by 10+ years of de-identified outcome records from schools across the Scoir network - Factors: GPA, test scores, high school characteristics (first-gen rate, geographic location), in-state/out-of-state status, round-level predictions - Scattergrams from both school-specific and network-wide data - College selectivity labels and match categories - Coalition app integration (alongside Common App) - Counselor document management and recommendation workflows
Pricing: Free for students and counselors. Colleges pay for enrollment marketing and recruitment analytics access.
Data sources: Tens of millions of de-identified application and outcome records submitted through Scoir by high schools since the platform's founding. Models retrained as admissions trends evolve.
Limitations: - Does not consider extracurricular activities, course rigor, or essays - Coverage depends on which high schools have adopted Scoir (network effects — may underrepresent certain regions or school types) - No hook modeling (athlete, legacy, donor) - No yield or enrollment decision simulation - School-level coverage varies
Differentiation from ABM: Scoir AI provides round-level individual probability estimates. An ABM runs hundreds of thousands of simultaneous agents through the full application-admit-yield-waitlist cycle, revealing systemic emergent patterns that no individual probability estimate can capture.
What it does: An all-in-one college planning platform (launched September 2023, consolidating Cappex, YouVisit, College Greenlight, and Match/Concourse) with a college acceptance calculator, scholarship search, and direct admissions features.
Target user: High school students in college search and application phase.
Key features: - College acceptance calculator: algorithm considers GPA, test scores, extracurriculars, demographic variables - Safety/match/reach classification - Direct Admissions: participating colleges proactively admit students based on Appily profile without application - Scholarship search - Virtual campus tours (inherited from YouVisit) - EAB's enrollment management data powering institutional analytics
Pricing: Free for students. Revenue from college enrollment marketing subscriptions.
Data sources: Average admitted student data points per college (Common Data Set, institutional data). Demographic variables included. No details on model validation or calibration.
Limitations: - Calculator methodology not publicly validated - No round-specific modeling - No hook, legacy, or donor modeling - The platform's commercial incentive (colleges pay for student leads) may bias school recommendations toward partner institutions - No yield or enrollment simulation
Differentiation from ABM: Appily is essentially a college marketing platform with a rudimentary chancing calculator. Its institutional client base (colleges) and student-facing tools serve enrollment marketing, not market simulation.
What it does: A college discovery and ranking platform built on user-generated reviews, aggregated public datasets, and student satisfaction surveys. Offers a college match quiz, acceptance probability calculator, and direct admissions.
Target user: High school students in the exploration phase of college search.
Key features: - "Your Best Fit" tool with user-adjustable sliders for academics, campus life, diversity, location, cost - Admissions calculator estimating acceptance probability by GPA, test scores, intended major - Side-by-side comparison tool (up to 4 colleges) - Niche Direct Admissions: real-time acceptances and upfront scholarships from partner colleges - Grades for colleges across categories (academics, professors, campus, diversity, value) - Scholarship search
Pricing: Free for students. Revenue from college advertising and lead generation.
Data sources: U.S. Department of Education federal datasets (IPEDS), survey data from students and alumni, Common Data Set. Grading methodology publicly documented.
Limitations: - Acceptance calculator is basic (GPA + test score inputs only) - Review-based rankings are susceptible to self-selection and gaming by institutions - No round-specific, hook, or EC modeling - Commercial incentives (partner colleges) affect recommendations - No dynamic competitive pool modeling
What it does: The College Board's free college planning platform offering college search, financial aid planning, and scholarship discovery across 4,000+ institutions. Tied to SAT score delivery.
Target user: All high school students, with particular reach into SAT test-takers.
Key features: - College search with 3,000+ profiles (college-provided content) - Net price calculators and financial aid planning tools - 30,000+ scholarships database - 2025 BigFuture School app for PSAT/SAT takers to receive scores and college match suggestions - No standalone chancing calculator (College Board avoids the prediction liability)
Pricing: Free.
Data sources: College-provided institutional data. College Board's SAT score database (>3 million test-takers per year). College Board Student Search Service for recruitment matching.
Limitations: - No chancing engine or admission probability model - College profiles are self-reported by institutions (marketing content) - Conflict of interest: College Board sells student data to colleges through Student Search Service - No competitive pool modeling, no yield or waitlist simulation
What it does: A transcript and credential exchange network used by 12,000+ high schools and 4,500+ colleges for secure document delivery. The institutional analytics side helps colleges understand their competitive position using aggregate transcript-flow and enrollment data.
Target user: Institutions (colleges and high schools). Also limited student-facing tools.
Key features (institutional): - Automated transcript data extraction and GPA/course count analysis for admissions processing - Competitive analytics: how an institution performs vs. peer institutions in applications, admits, and yields - Student Choice Rankings: ranks colleges by actual enrollment decisions when students were admitted to multiple schools (based on tens of thousands of observed pairs) - Recruit module: analytics to identify high schools with untapped recruitment potential - Yield rate: institutions using Parchment Recruit average 39% yield
Pricing: Institutional SaaS licensing (not publicly disclosed). Student-facing transcript sending is per-document fee.
Data sources: Actual transcript delivery data across the Parchment network — one of the highest-quality enrollment outcome datasets in existence, because it captures actual matriculation rather than self-reported outcomes.
Limitations: - Primarily an institutional tool; not student-facing - Student Choice Rankings are aggregate, not predictive for individual students - No simulation capability
Differentiation from ABM: Parchment has exceptional real outcome data but uses it only for institutional benchmarking and recruitment marketing, not for simulating market dynamics or providing student-facing probability estimates.
What it does: Common App's member analytics platform provides benchmarking data to the 1,000+ member colleges about application volume, demographics, submission trends, and peer institution comparisons.
Target user: College admissions offices and institutional research staff.
Key features: - Peer benchmarking: compare your institution's application trends vs. anonymized peer groups (groups of 10+ institutions, aggregated for privacy) - Application trend data: volume by demographics, geography, test score submission rates - Research briefs and state-level reports - Data analytics and research team publishes annual Common App reports
Pricing: Included in Common App member institutional fees (not disclosed separately).
Data sources: The Common App application platform itself — actual application data from all ~1 million+ annual applicants to 1,000+ member colleges. Among the highest-quality application datasets in existence.
Limitations: - Not accessible to students or counselors directly - Aggregate only; no individual prediction capability - No simulation or modeling of yield dynamics
Several tools implement the Ivy League's Academic Index formula, which was developed in 1985 to ensure recruited athletes meet minimum academic standards for Ivy League schools. The formula produces a score on a 60–240 scale.
Formula (classic): (1/3 × SAT composite) + (1/3 × SAT Subject Tests average) + (1/3 × Converted Rank Score), each component scaled to 80.
Updated formula (post-2016 score recentering): Adds EBRW + Math score + Converted GPA Score (CGS).
Tools: - ivyleagueguru.com/calculate-index: Inputs SAT/ACT + class rank/GPA; outputs score and band (below 171 = not recruitable; 171+ = High/Medium/Low/Low-Low bands). Free. - toptieradmissions.com Academic Index Calculator: One of the earliest web implementations. Free, lead generation for consulting services. - academicindex.ai: Extends the AI concept beyond athletics to general holistic admissions. Free. - calculator.academy and calculatorultra.com: Generic calculator implementations. Free.
Limitations across all AI calculators: - Most do not account for the 2023 digital SAT scoring changes - None model test-optional applicants - The original AI formula was designed specifically for athlete recruiting floors, not general admissions - No per-college probability output; just a single score
What it does: A free AI-powered college admissions predictor using ensemble ML (Random Forest + XGBoost) trained on IPEDS federal data and Common Data Set reports from 300+ universities (2018–2024).
Target user: High school students, particularly first-generation and underserved applicants (explicit mission statement).
Key features: - 20-factor holistic profile analysis - Coverage: 1,600+ colleges in detailed model; 2,100+ accredited U.S. institutions with some coverage - Essay scoring tool with admissions rubric - Real-time profile updates as student improves profile - Community insights from successful applicants
Pricing: 100% free; explicitly committed to equal access mission.
Data sources: IPEDS federal data (DOE), Common Data Set reports (2018–2024), College Scorecard data, anonymized student outcomes. Calibrated with isotonic regression; Brier score validation.
Limitations: - Historical data may not capture year-over-year admissions variation - 1,600 colleges modeled — strongest coverage at mid-selectivity; Ivy/HYPSM predictions rely on thin data - No round-specific (ED/EA/RD) modeling disclosed - No hook, legacy, or athlete modeling - No competitive pool or yield simulation
Several academic papers have modeled college admissions as an agent-based market. These are the direct intellectual predecessors to what the college-sim project is building.
Reardon, Kasman, Klasik, Baker (2016) — JASSS - "Agent-Based Simulation Models of the College Sorting Process" - Two agent types: Students (resources + caliber) and Colleges (quality = avg enrolled caliber) - Three-stage annual loop: Application → Admission → Enrollment - Five resource pathways examined: achievement gap, application enhancement, information access, application volume, utility valuation - Key finding: resource-caliber correlation explains ~60% of enrollment stratification; other four pathways combined explain roughly the same amount again - Policy extensions: examined race- and SES-based affirmative action scenarios - Platform: academic working paper; not a consumer product
Assayed & Maheshwari (2024) — SSRN / CSEIJ - "A Review of Agent-Based Simulation for University Students Admission" - Survey of ABM applications in university admissions internationally - Classifies models by: educational attainment level, university selection behavior, socioeconomic linkage to college choice and financial aid - Finds NetLogo-based implementations with family-income and GPA parameters optimizing fairness/equality utilities - Key finding: peer influence is a significant determinant — juniors more likely to enroll if peers with similar SES background succeeded academically
Gap: All academic ABMs are policy research tools with synthetic data, not real-data-driven consumer products. None use actual U.S. college statistics (acceptance rates, SAT ranges, ED multipliers, yield data) calibrated to real institutions.
| Feature | academicindex.ai | CollegeVine | Scoir AI | Naviance | Appily | Chancify AI | Parchment | Academic ABMs | College-Sim (this project) |
|---|---|---|---|---|---|---|---|---|---|
| Per-college admit probability | Partial (bands) | Yes | Yes | Via scattergrams | Yes | Yes | No | Synthetic | Yes (logistic model) |
| ED/EA/RD round modeling | No | Partial | Yes | No | No | No | No | No | Yes (6 rounds) |
| Hook modeling (athlete, legacy, donor) | No | No | No | No | No | No | No | No | Yes (3.5x–4x multipliers) |
| Extracurricular tier modeling | Partial | Yes (12 tiers) | No | No | Partial | Partial | No | No | Yes (EC/essay component) |
| Population-level simulation | No | No | No | No | No | No | No | Yes | Yes (ABM) |
| Yield & enrollment decisions | No | No | No | No | No | No | No | Partial | Yes (utility model) |
| Waitlist dynamics | No | No | No | No | No | No | No | No | Yes (fill-rate trigger) |
| High school feeder effects | No | Partial (geography) | Yes | Yes (school-level) | No | No | Yes | No | Yes (20 HS archetypes) |
| Income/SES modeling | No | No | No | No | No | No | No | Yes | Yes (Chetty yield offsets) |
| Race/demographic modeling | No | Partial | No | No | No | No | No | Yes | Yes (post-SFFA bars) |
| Real college data (30 colleges) | No | Yes | Yes | Yes | Yes | Yes | Yes | No (synthetic) | Yes (30 real colleges) |
| Free to use | Yes | Freemium | Free | School license | Free | Yes | Institutional | Academic | Yes (open HTML) |
| No server required | N/A | No | No | No | No | No | No | No | Yes (single HTML file) |
| Counselor / institutional focus | No | No | Yes | Yes | Partial | No | Yes | No | No (currently student-facing) |
| Simulation replay / scenario testing | No | Partial (what-if) | No | No | No | No | No | Yes | Yes |
Every competitor computes P(admit | student). None compute "given 50,000 students applying this cycle, what fraction with profile X will actually enroll?" The difference is not academic — ED fill rates, waitlist movement, and yield cascades are emergent phenomena that only appear when you simulate the whole market. An ABM is the only architecture that captures this.
Admissions is not a single probability computation — it is a six-round sequential matching market (ED → EA/REA → EDII → RD → Decisions → Waitlist). Each round changes the pool available for the next. No consumer tool models this sequence; most compute a single probability and stop.
The athlete (3.5x), donor (4x), legacy (2.5x), and first-gen (1.4x) hook multipliers are real, documented, and large. CollegeVine acknowledges it does not model these. Scoir AI does not model them. No free tool models them. The college-sim project has explicit hook modeling in the logistic admissions formula.
Colleges are not passive admission machines — they have yield targets, class composition goals, financial aid budgets, and reputational incentives. The college-sim's college-as-agent architecture captures fill-rate management and waitlist triggering. No competitor models the college side of the market.
Because the ABM simulates the entire market, it can answer questions that no individual chancing tool can: "What happens to UCLA's yield if Harvard expands its class by 10%?" or "What is the effect of eliminating the donor hook on socioeconomic diversity at HYPSM?" This is the Reardon et al. (2016) use case — and it has never been productized.
All commercial competitors are web services requiring server infrastructure, authentication, and ongoing data pipeline maintenance. The college-sim is a single self-contained HTML file with real data embedded in JSON. It can run offline, be shared as a file, be embedded in a course, or be forked for research. This is a uniquely accessible architecture.
High school counselors currently use Naviance scattergrams (school-specific, small N) or Scoir AI (network-wide, no EC/hook). Neither shows how a counselor's entire 12th-grade cohort will sort across the admissions landscape simultaneously. A counselor-facing ABM that simulates "here is how this year's senior class will likely distribute" is a genuinely novel tool with no existing competitor.
Since the Supreme Court's 2023 Students for Fair Admissions decision, colleges have been scrambling to understand how to maintain diversity without explicit race-conscious admissions. The Reardon (2016) model analyzed affirmative action scenarios. A real-data-calibrated ABM could simulate the enrollment diversity impact of specific policy changes (e.g., eliminating legacy hooks, expanding first-gen multipliers, geographic diversity weighting) — a tool with clear value for institutional research teams.
No consumer tool models the full college decision, which includes not just "will I get in" but "what will it cost me." The Chetty yield data and net-cost-by-income data in the college-sim project are exactly the inputs needed to model "which school will I actually choose given my family's income and the financial aid offers I receive." The intersection of admission probability + financial aid offer + yield is uncovered territory.
The U.S. college admissions market receives ~1 million international applications annually. International applicants face different effective acceptance rates, have no access to Naviance scattergrams, and are largely ignored by tools calibrated on domestic applicant pools. An ABM with international applicant archetypes would serve this market exclusively.
Transfer students represent ~40% of U.S. undergraduates. Every tool in this landscape focuses on first-year, direct-from-high-school admissions. The transfer admissions market — with different GPAs, articulation agreements, and college-level course credits — has no dedicated simulation or chancing tool.
Naviance is a $10–$30/student/year school district tool. Its scattergrams are its primary differentiator, but they degrade for small schools. An ABM-powered counselor tool that supplements sparse school-level data with national simulation could command similar licensing fees while providing meaningfully better insights. The admissions consulting market is $2.3–3.4 billion; the institutional EdTech market that Naviance serves is a separate, durable revenue stream.