Questions and answers.

Only the teacher. School leaders and the local authority see whole-school patterns, never a lesson, a reflection, what a teacher chose to try, or who has recorded a lesson or how often. They do see who has joined their school (names, work emails and join dates), so they can manage access, and how many staff chose each indicator as a focus when recording, once five or more have.

Yes, at whole-school level. Leaders compare terms, school years and short sprints of up to two weeks, set to the day by the local authority’s term dates, on screen or as an Excel file or printed summary. A period shows figures only when at least five different staff recorded in it, and in every part of it, so comparing periods can’t single out one teacher. A change shows what the recordings captured; the staff who chose to record may differ from one period to the next. Staff hear about a sprint through a card on their record page and a staffroom poster. Taking part is their choice, and nobody sees who did.

No. Teachers choose which lessons to record, nothing is scored, and no one else can read a reflection. Whole-school figures are a share of lessons, never of staff, and appear only once five staff have joined.

The audio is deleted within seconds of being transcribed, and the transcript as soon as the lesson has been read. The quoted moments are removed after 14 days, or up to six weeks if the teacher starts a reflective cycle from that lesson.

Names are replaced with labels such as [Pupil A] as soon as the lesson has been turned into text, before it is saved or sent for reading, so the reading and the moments you get back use labels, not names. It reduces the risk but does not remove it: a name the software does not recognise can be missed, and a detail that identifies a child without naming them is not touched. The audio is deleted within seconds.

Your lessons never leave the UK. Lessons, reflections, cycles and accounts are stored in the UK, and the service runs in London. To transcribe and read a lesson, it is sent to Microsoft and Amazon Web Services under data processing agreements. Microsoft receives the audio and turns it into text using its Azure AI Speech service in the UK (UK South). Microsoft keeps none of it, and processes it only to provide the transcription; we delete our copy as soon as the text is back. Amazon receives only the text, with pupil and staff names already replaced by labels, and processes it in London. Amazon keeps none of it: we have switched on Amazon Bedrock’s zero data retention setting. During the pilot, sign-in codes are emailed by Brevo, in the European Economic Area, which keeps a log of each email, including the address it went to, for one month.

Every provider is listed in our privacy notice.

Second Listen is licensed to a whole local authority, covering every school and teacher in it. Pricing details to follow; get in touch in the meantime.

With their own work email: we send a six-digit code, so there are no passwords. They join their school once, with a school code or a QR poster from the staffroom.