Teachers: The missing link in personalised learning

Personalisation: this word has come up with increasing frequency in conversations with school leaders over the past year, one word has come up with increasing frequency. The discussion often moves quickly to technology. AI tutors, adaptive platforms, and real-time dashboards promise to help schools understand what every child needs and respond accordingly. Perhaps, the bigger question is whether the question itself is being asked in the right order.Imagine a hospital that installs the world’s finest diagnostic machines and then cancels the training budget for its doctors. The scanners hum, the dashboards glow, every ailment is flagged in real time but no patient is treated better than before. Nobody would call that personalised medicine. Yet that, to a large extent, is the current situation in the education ecosystem. For years, we have been investing in devices, adaptive software, and AI tutors. All of it is based on the quiet assumption that personalisation is something machines do.It is not. Personalised learning is not a software feature but a teaching skill. Until the momentum behind education technology shifts from acquiring hardware to building teacher capability, the personalisation promise will stop at the classroom door. To be fair, technology has earned its place. Adaptive platforms have genuinely changed what is possible: checks for understanding that once took a week of correction now take a minute. A teacher can see, mid-lesson, exactly which children have mastered a concept and which have not. Nor is this a bubble. India’s ed-tech market, valued at $7.5 billion in 2024, is projected to reach $29 billion by 2030, according to Impact Study of EdTech in India: Driving Innovation and Creating Opportunities, IAMAI and Grant Thornton Bharat, January 2025. The tools are real, and so is the money behind them.Where teaching beginsBut consider what a tool actually knows. A dashboard can flag a child struggling with fractions but cannot know why. Did she misread the word problem, did not understand the concept, or has she simply learned to fear being wrong in front of others? Each reason demands a different response and only one person who can diagnose and act on it is the teacher. What happens in the 10 minutes after the flag determines whether personalised learning is a reality or a slogan? No software controls those 10 minutes.The question, then, is whether teachers have the knowledge and conditions to do something with it. That is where personalised learning is won or lost.First, data is not insight. Learning data becomes learning only when a teacher converts it into timely classroom action: a concept re-taught differently the next morning, a small group pulled aside that same week, a task adjusted for the child who is ready to move ahead. Without that conversion, a dashboard is just a more expensive way of discovering, too late, what a good teacher would have noticed anyway. The last mile of personalisation is human.Second, capability is the multiplier. Place identical software in two classrooms and the results diverge entirely. In one, the data drives regrouping and re-teaching; in the other, it decorates a monthly report. The variable is never the licence fee. It is the teacher’s confidence in reading evidence and responding to it. This means the returns on developing teachers now exceed the returns on buying machines.Third, that capability is built continuously, not ceremonially. A one-off workshop in June does not produce a data-confident teacher in November. What does is unglamorous and proven: teachers meeting weekly in structured professional learning communities to examine student work and evidence together; short plan-do-study-act cycles that test one instructional change at a time; coaching inside the classroom rather than lectures outside it; and self-paced online modules that let a teacher learn on her own schedule. None of this requires new devices. All require schools to treat teacher time as an investment rather than a cost.Fourth, the classroom itself must flex. Personalisation dies in a room where all 40 children do the same worksheet at the same pace. It lives in flexible grouping that reshuffles as the data changes, tiered support that gives struggling learners more time and attention without stigma, and differentiated tasks that stretch every band of the class. These are pedagogical moves, not product categories and no app performs them.Beyond the metrosPerhaps the most important consequence of this argument is that personalised learning does not have to be a privilege of affluent schools. The prevailing assumption is that it must trickle down from elite urban schools once budgets allow. The opposite is true. The teacher-capability route is cheaper than the hardware route and it travels better. A smart classroom needs a capital budget; a professional learning community needs a timetable slot and a protocol. Well-designed online professional development reaches a teacher in Nashik as easily as one in Mumbai. Personalisation beyond the metros is not a procurement problem. It is a capability problem, which is why it is solvable.The roadmap, then, is short. Audit what is spent on technology against what is spent on the people using it, and correct the ratio. Protect weekly collaborative time for teachers as fiercely as instructional time for students. Make data review a standing routine that always ends in a classroom action and not in a filed report. Measure professional development by changed teaching practice, not certificates issued.The diagnostic machine never cured anyone; the doctor did. The most advanced personalisation technology in any classroom is, and will remain, the teacher.The writer is co-founder and chairperson, The Acres Foundation.