From Guesswork to Data: How Transcript-Based Lesson Analysis Is Changing Teacher Coaching

For as long as schools have run coaching programs, lesson feedback has relied on the memory and impressions of an observer. Even the most skilled coach can only see and hear so much in real time — and what gets remembered often depends on what stood out most in the moment. The result? Teachers walk away with valuable but selective feedback, while countless micro-moments of learning vanish into thin air.

Transcript-Based Lesson Analysis (TBLA) changes this. Born out of lesson study traditions in Japan and now adopted in countries from Indonesia to the US, TBLA takes a simple idea — turning classroom talk into text — and makes it the foundation for rich, evidence-based coaching. A transcript preserves every question, every student contribution, every pause. It’s a mirror that never blinks.

Why TBLA Works

When teachers review transcripts together, they can:

• Measure talk time: Exactly how much did students speak compared to the teacher?

• Spot questioning patterns: Were prompts open-ended or closed? Did they encourage reasoning?

• Track student-to-student dialogue: Was discussion truly collaborative, or did every exchange route through the teacher?

• Examine feedback quality: Were comments specific and constructive, or simply evaluative?

In study after study, TBLA has revealed hidden factors that shape learning — from over-reliance on low-quality questions to the absence of peer-to-peer interaction. These are patterns that test scores and traditional observations often miss.

Adding AI to the Mix

Until recently, TBLA was a time-consuming process. Recording, transcribing, and coding a single lesson could take hours. Now, AI makes it scalable:

• Automated transcription turns lesson audio into text in minutes.

• Natural Language Processing detects key teaching moves, from probing questions to praise.

• Sentiment analysis flags shifts in classroom climate.

• Dashboards transform raw transcripts into clear, actionable insights.

In the same time it once took to type out a paragraph, teachers can now receive a breakdown of talk ratios, examples of strong questioning, and reflection prompts tailored to their lesson.

Why This Matters for UK Schools

With CPD budgets under pressure, schools need tools that deliver high-impact feedback without the logistical overhead of repeated in-person observations. AI-enhanced TBLA offers:

• Scalability: Every teacher can receive regular feedback, not just those observed in a formal cycle.

• Consistency: Insights are drawn from the full record, not selective memory.

• Professional trust: Data is private to the teacher, supporting growth without judgement.

It’s not about replacing human coaching — it’s about giving coaches and teachers richer evidence to work with.

The Next Step

This September, Starlight will bring AI-enhanced TBLA into everyday school life, turning lesson audio into specific, timely, actionable and regular feedback in just minutes. The goal is simple: make high-quality coaching available to every teacher, every week, without adding to workload.

The age of guesswork is over. With the right tools, we can ground professional growth in evidence — and give every teacher the feedback they deserve.

The Insight Engine is written by Adam Sturdee, co-founder of Starlight—the UK’s first AI-powered coaching platform—and Assistant Headteacher at St Augustine’s Catholic College. This blog is part of a wider mission to support educators through meaningful reflection, not performance metrics. It documents the journey of building Starlight from the ground up, and explores how AI, when shaped with care, can reduce workload, surface insight, and help teachers think more deeply about their practice. Rooted in the belief that growth should be private, professional, and purposeful, The Insight Engine offers ideas and stories that put insight—not judgment—at the centre of development.