Home › Data & Findings
Key metrics, visualised
The charts below present the quantitative results exactly as reported. Baseline figures are shown in slate grey and follow-up figures in teal.
Weekly time spent on lesson planning (hours)
A 52.5% reduction
Average weekly planning time fell from 10 hours to 4.75 hours — a 52.5% reduction, or about 5.25 hours returned to each teacher every week. The gain was consistent across measures: median planning time dropped from 7 to 4 hours (42.9%) and maximum time from 24 to 10 hours (58.3%). This is the report's single most striking quantitative finding, and the foundation for every wellbeing and quality gain that follows.
Source: teacher baseline (Nov 2024) and follow-up (Feb 2025) surveys.
Planning efficiency, confidence and fit to class needs (%)
Quality rose with efficiency
Time savings did not come at the expense of quality. Staff rating their planning efficient or very efficient rose from 0% to 88%. Teacher planning confidence rose from 50% to 100%, and confidence that planning consistently meets class needs rose from 37.5% to 100%. The adaptability of plans for SEND and EAL learners rose from 50% to 88% — AI made planning both faster and more inclusive at once.
Source: teacher baseline and follow-up surveys.
How lesson planning affected workload & wellbeing (% of staff)
The burden flipped
The picture inverted completely. Staff seeing planning as a high burden on workload fell from 75% to 0%, while those seeing it as a low burden rose from 0% to 75%. Feeling often or always overwhelmed by planning fell from 75% to 0%. The report adds important nuance: although the experience of planning stress collapsed, teachers frequently found the time saved was redirected to other tasks rather than reducing total working hours — so leaders must actively protect these gains.
Source: teacher baseline (Nov 2024) and follow-up (Feb 2025) surveys. Every bar is labelled, including values that fell to 0%.
Reported impact of AI on inclusion (% reporting each outcome)
Consistent inclusion gains
These are follow-up figures — the proportion of staff and pupils reporting each outcome after AI adoption (there is no baseline for these particular survey items). They were strong across the board: 95% said AI improved the speed and quality of differentiated materials, 91% of KS2 pupils preferred AI-supported lessons, 89% said AI gave more equitable access "without lowering the bar", 82% reported increased pupil engagement and 78% improved learner independence.
Source: follow-up staff surveys and pupil voice (Spring 2025). These items were collected at follow-up only, so they show the share reporting each outcome rather than a before/after change.
All headline metrics
The complete set of quantitative findings, for reference and as a text alternative to the charts above.
| Measure | Baseline | Follow-up | Change |
|---|---|---|---|
| Average weekly planning time | 10 hrs | 4.75 hrs | −52.5% |
| Median weekly planning time | 7 hrs | 4 hrs | −42.9% |
| Maximum weekly planning time | 24 hrs | 10 hrs | −58.3% |
| Rated planning efficient / very efficient | 0% | 88% | +88 pts |
| Teacher confident / highly confident in planning | 50% | 100% | +50 pts |
| Confidence planning consistently meets class needs | 37.5% | 100% | +62.5 pts |
| Plans adaptable for SEND / EAL learners | 50% | 88% | +38 pts |
| Planning has a "high impact" on workload | 75% | 0% | −75 pts |
| Often / always overwhelmed by planning time | 75% | 0% | −75 pts |
| Increased pupil engagement (teacher-reported) | — | 82% | 82% |
| Improved learner independence (teacher-reported) | — | 78% | 78% |
| KS2 pupils preferring AI-supported lessons | — | 91% | 91% |
| More equitable access "without lowering the bar" | — | 89% | 89% |
| Improved speed & quality of differentiated materials | — | 95% | 95% |
Reading the data together
The numbers tell a coherent story: a large, consistent fall in planning time and stress, matched by a rise in confidence, quality and inclusion — with no measure moving in the wrong direction. The one figure a leader should not read at face value is the wellbeing gain. Because saved time was often reabsorbed into other work, the collapse in "overwhelm" reflects a better planning experience rather than a guaranteed reduction in total hours. The data is a strong case for AI-supported planning — and an equally strong case for actively protecting the time it frees.