Winter feedback: understanding how teachers are using AI marking
Haris Mughal is the Head of Growth at Running Paper.
The evaluation
AI supported marking is no longer something schools are approaching cautiously. Over the past year, it has moved from something teachers were curious about to something they are actively building into their assessment workflows. But adoption alone doesn't tell you much. The question we kept coming back to was simpler and more important: once teachers start using it, do they keep using it?
To find out properly, we ran a structured evaluation with 53 teachers across 13 secondary schools during the Autumn term. Teachers used Running Paper for formative assessments, end of topic tests, and mock examinations across a range of subjects. Real classroom use, over a full term.
What stood out most was sustained use. Teachers were not trying it once and returning to their old process. They continued using it across multiple assessments as part of their normal workflow.
The results
Key statistics:
- 87%: would recommend Running Paper to a colleague
- 83%: said feedback helped improve student understanding
- 2.4 hrs: saved per class on formative assessments, on average
- 3.1 hrs: saved per class on mocks and summative work
81% of teachers said question level analysis helped them identify areas to focus on in future lessons. Across a full term, across multiple classes, that is a significant reduction in one of the most time-intensive parts of the job.
Beyond saving time
Time savings were expected. What teachers kept coming back to was something slightly different: how quickly they could understand their class.
Normally, knowing how a group of students performed across a set of questions means working through every script manually before the picture becomes clear. With Running Paper, that insight is immediately available. Patterns across the class, misconceptions that need addressing, areas to revisit in future lessons. Assessment insight that used to take hours to compile was simply there.
Instead of spending hours marking and compiling reports, teachers could identify where their class had struggled and plan accordingly, before the next lesson.
Built on teacher feedback
Running Paper is a teacher-first platform. A large proportion of the improvements made over the past year have come directly from classroom use and teacher responses. This includes refining marking moderation and confidence metrics, making reports clearer and more actionable, linking assessment objectives to individual questions, and improving workflows so the platform fits naturally into existing assessment processes.
The aim has never been to change how teachers assess. It is to remove the most repetitive, time-consuming parts of the process while improving access to meaningful insight, so teachers can spend their time on the work that actually requires their expertise. Ongoing development will continue to be shaped directly by teacher feedback.
What this tells us
Sustained use, strong recommendation rates, and measurable impact on both workload and classroom insight provide clear evidence that AI supported marking is already delivering real value in real schools. This evaluation reflects a broader shift, one driven not by hype, but by teachers finding genuine, practical value and choosing to keep coming back.