For principals · counselors · superintendents
Your teachers get hours back. You get a number for the board.
It runs inside the lesson they already planned, so there is no new block to timetable and nothing to mark. Cohesion is measured privately, class by class, term after term.
What the pilot found
Teachers gained hours. The rooms that ran it changed more.
Half the classes at Svenstrup School in Aalborg, Denmark ran it for six weeks; half did not. Cohesion and engagement rose about ten times more in the classes that ran it, and unexcused absence moved in opposite directions. Attendance is the one number nobody can talk themselves into — and it is a Danish school, not randomised, so treat it as a signal and take your own reading in a pilot.
Where that sits in ESSA’s tiers, since you will be asked: Tier 4 — demonstrates a rationale. The design is quasi-experimental, which is Tier 2 territory, but Tier 2 also wants two sites and this was one school. We will not write a tier we cannot defend, and you should ask every vendor in the room for their sample size and their site count.
| Six weeks, 800 students | VEN classes | Control |
|---|---|---|
| Class cohesion3.1 → 4.0 on a five-point scale · Cohen’s d = 1.42 | +29% | +3.1% |
| Student engagement3.0 → 3.9 · Cohen’s d = 1.38 | +30% | +3.3% |
| Unexcused absenceunexcused absences, school records | −18% | +2% |
| Teacher time savedself-reported, with daily use | 4.3 h/week | — |
Svenstrup School, Aalborg, Denmark, 21 October – 1 December 2025. 800 students, 40 teachers, controlled, stepped design (two intervention cohorts, two comparison cohorts); classes not randomised. Cohesion and engagement are student self-report on validated five-item scales (α = 0.89 / 0.85); absence is from school records. One pilot, published in full with its limitations: read the study (PDF) · what it does and doesn’t show.
Built the way a school would build it
One database per school
Your school’s data is physically separate from every other school’s. Nightly backups. Export everything, any time, in CSV or JSON. Deletion on request.
No student accounts
Students never log in, never install anything, never give an email. The system holds a first name or nickname per child and nothing else that identifies them. COPPA and FERPA obligations stay small because the data does.
No monitoring, by design
No cameras, microphones, content scanning, risk scores or alerts. Everything the system knows, it knows because a person tapped something. Safeguarding stays with your people and your procedure.
Teacher-approved, always
Questions are drafted before class, checked by a second model against the published protocol, and approved one by one by the teacher. There is no approve-all, and nothing that writes questions has a surface a child can reach.
Rostering that fits
CSV from your SIS, OneRoster sync, or names typed in. Google Classroom, Clever and ClassLink connections are on the roadmap and will be listed here only when they work.
Readable from the back row
Type sized to the viewing distance, every colour named in words, keyboard operable, reduced motion respected. Stage in English and Danish; questions pre-translated into two more languages per session.
Pricing
The method is free. The hours are what you buy.
The method
Free, in full, forever.
$0 CC BY-SA 4.0
- The whole protocol: the round, the safety rules, how the questions are written.
- Run it tomorrow with four coloured cards and a printed sheet.
- Adapt it, translate it, teach it, build your own. No permission needed.
The software · schools and districts
The hours, and the record.
$1 per student per month
- Questions written for this week’s topic — connection and academic — ready to approve in two minutes.
- The Stage, the pairing, the languages, the private check-out.
- Class cohesion across terms and years, per class and per competency.
- A separate database per school, CSV or OneRoster rostering, exports and deletion on request.
- A quote on request, invoicing against a signed service agreement with a data-privacy clause, and your state’s student-data privacy agreement signed where one exists.
- Updates and support from the founder, not a queue.
Billed annually by invoice against a signed service agreement. Districts: the same price, one agreement. No setup fee and no training fee — the method is published, and the app explains itself.
Questions school leaders ask
What does it cost?
$1 per student per month, every feature included, billed by invoice against a signed service agreement. No procurement portal and no per-seat surprises. The method itself is published free under an open licence — what you are buying is the software, the record across terms, and the hours it gives your teachers back.
Where does our data live?
Each school gets its own separate database on Cloudflare’s network; nothing is pooled across schools. The system stores a first name or nickname per student, the class’s questions, who paired with whom, timings, the teacher’s private ratings and the 1–5 check-out answers. No emails, birth dates, photos, audio or video of students, ever.
How does rostering work?
CSV import from your SIS export, a OneRoster sync, or teachers adding names by hand. Student caps and teacher caps are enforced server-side on every path.
What actually writes the questions?
It drafts questions before class from the grade band, the focus and the topic — never from a child’s name or record — and a second model checks each draft against the protocol’s rules before a teacher ever sees it. The teacher approves, edits or discards every one. Nothing that writes questions has an interface a student can reach, and if it is unavailable the curated bank fills in. The full disclosure, including which model and where the request goes, is on the privacy page.
Does it monitor students?
No. No cameras, no microphones, no content scanning, no emotion recognition, no risk scores, no alerts. Safeguarding follows your existing procedure with your existing people. The absence of sensors is written into the patent as a claim, not a setting.
Is it accessible?
The Stage is built for the back row: type sized to the viewing distance, high contrast, no reliance on colour alone (every colour is named in words), keyboard operable, reduced motion respected. The public pages and the app target WCAG 2.1 AA; ask for the current accessibility statement.
What about FERPA and COPPA?
There are no student accounts and no student contact details, so the surface area is small by design. Schools keep control of their data, can export it in full, and can have it deleted on request. A service agreement with a data-privacy clause is part of every paid setup; we will sign your state’s student-data privacy agreement where one exists.
Start with one class.
A pilot needs no procurement. If the room changes, you will know inside three weeks, and you will not need us to tell you.