Education · Sentiment Analysis · AISentiment Analysis for a PE-backed Private School Network
A proof of concept applying LLM-powered sentiment analysis to student reviews across 8 campuses, surfacing operational gaps and value creation levers for a PE-backed education group.
Online reviews are an underexploited data asset for PE-backed education operators. This proof of concept by REKOLT demonstrates how publicly available student feedback, when structured through an LLM pipeline, can generate actionable intelligence on campus performance, brand consistency, and satisfaction drivers. Conducted across two schools and eight campuses of a French private education group, the analysis covers verified student reviews from 2019 to 2024, structured along five dimensions: teaching, facilities, atmosphere, internships, and value for money.
Key findings:
Atmosphere, not academics, drives overall satisfaction,making perception harder to control and underlining the limits of purely operational improvement plans.
Weak franchise effects. Student experience varies significantly across campuses of the same brand, revealing an absence of standardization that directly challenges scalability under a value creation plan.
Satisfaction is declining network-wide. Both brands have seen steady rating drops since 2021–2022, potentially signaling structural issues beyond external factors like COVID or job market volatility.
Nearly 1 in 10 students explicitly rates the experience as not worth the cost, a small but reputationally dangerous signal for premium-priced institutions.
Each campus requires a localized strategy. No single lever performs equally across sites, making a one-size-fits-all transformation plan insufficient.
This study is a proof of concept. Its full value would emerge when cross-referenced with internal operational and financial data, and benchmarked against peer institutions, exactly where REKOLT's data and AI capabilities come in.
