37 % fewer drop-outs and an AZAV audit that no longer ruins the weeks before it.
Detailed case study: AZAV-certified training provider, 240 participants in 9 parallel cohorts, publicly funded. Superchat as the central channel per cohort with drop-out triage and an audit-proof communication log.
Anonymised case study
This is an anonymised case study based on typical project parameters from comparable Superchat implementations. Numbers, quotes and details are constructed plausibly to reflect realistic outcomes; no real customer data is shown. Real case studies with logos and original numbers will follow once we have written approval from the respective companies.
Profile: AZAV-certified training provider, 240 participants, 22 staff
Starting point
The provider operates three sites and runs publicly funded courses for career returners, career changers and participants from the German SGB-II/III welfare context. Each site runs 2–4 cohorts in parallel, each between 8 and 20 weeks long.
Communication with participants used to run via the private WhatsApp accounts of the respective course leaders. Fast, but legally borderline, intransparent, and resulting in total knowledge loss whenever a course leader changed. There was no central documentation of who communicated with whom about what and when.
The drop-out rate during the first four weeks of a course was 29 %. Course leaders typically only spotted drop-out risk once a participant had already missed two days unexcused — at which point re-engagement was statistically nearly hopeless. For the employment agency as funding body, the drop-out rate is a central steering KPI; every premature drop-out costs the provider reputation and potential future referrals.
AZAV documentation for the annual external audit was maintained manually in Excel sheets. Per audit, the quality management department spent roughly three person-weeks on assembly, plausibility checks and chasing missing evidence.
Goals
- A GDPR-compliant communication channel per cohort that survives staff changes
- Early-warning system for drop-out risk in the first 30 days of each cohort
- Structured module progress updates so participants can catch up on their own
- Auto-generated AZAV audit report in under two hours, complete and plausible
- +25 % referrals from the employment agency through better performance reporting
Approach
Phase 1 — Setup & legal framework
Week 1Superchat Advanced provisioned, one dedicated WhatsApp business number per cohort (instead of one provider number for everything), DPA plus a specifically worded consent clause for participants. The clause was reviewed with the in-house data protection officer.
- 9 verified WhatsApp business numbers (one per active cohort)
- DPA + privacy consent as mandatory annex to the funding application
- Course leaders' private mobile is explicitly no longer in use
- Role model: course leader read/write, social pedagogy read-along, QM audit-only
Phase 2 — Drop-out early warning
Weeks 2–3Automated check-ins at fixed points during the first half of each course. Reply, non-reply and reply content feed a simple risk heuristic. On a red status, social pedagogy receives a task in the internal tool to make personal contact.
- Day-3 check-in: "How was your start?" (5 buttons: great → tough)
- Check-ins on day 7, 14, 30: module-specific short questions
- Trigger: non-reply > 36 h or "tough" answer → social pedagogy is alerted
- Re-engagement rate after intervention on red status: 64 %
Phase 3 — Module updates & audit export
Weeks 4–5Weekly rhythm for module progress updates, automatic dispatch of learning material links, deadline reminders for internships and exams. A clean audit export was designed jointly with QM and validated against an old audit dataset.
- Per module a mini-plan as WhatsApp message (duration, learning goals, material link)
- Deadline reminders T−7 / T−2 / T−0 for exams and internship registration
- Audit export: complete communication log of a cohort as PDF + machine CSV
- QM validation of the export against a historical audit (all required fields present)
Concrete flows and templates
Day-1 onboarding flow
The first day of every cohort opens with a structured greeting sequence: course leader introduces themselves, house rules as PDF, schedule link, emergency contact options.
Drop-out early warning
Three automatic check-ins in the first 14 days. Replies are scored with a simple traffic-light logic. Red statuses are followed up by social pedagogy within 4 h.
Weekly module update
Every Sunday at 6 pm a brief weekly preview goes out: modules of the week, learning goals, self-study material, dates.
AZAV audit export
For a chosen cohort and time period: a PDF report (auditor-readable) plus a CSV (machine-readable) covering all communication touchpoints, consent records and re-engagement actions.
Results (before go-live → 6 months after go-live)
| KPI | Before | After | Delta |
|---|---|---|---|
| Drop-out rate in first 4 weeks | 29 % | 18 % | −37 % |
| Re-engagement rate on red status | not measurable | 64 % | newly unlocked |
| Employment agency referrals (quarter) | ~ 95 | ~ 120 | +26 % |
| AZAV audit prep effort | ~ 3 person-weeks | < 2 hours weekday | −98 % |
| Completeness of mandatory records | ~ 84 % | 100 % | audit-proof |
| Course leader response time | unclear | < 30 min during day | transparent |
Voices from the training provider
„Our funding agency called us the most organised training provider in the region. A year ago I would have laughed at that sentence."
„I now see on day 3 if someone is wobbling — and can step in deliberately. We used to only notice when someone had stopped showing up."
„The early warning kept me in the course. The social worker called me when I almost quit after a tough week."
Lessons learned
- One business number per cohort (instead of one provider number) was extra effort upfront, but paid off massively: no context switching, clear ownership, simpler audit export. The added verification effort was negligible.
- Drop-out early warning only works in combination with quick social-pedagogical intervention — the technology alone would not have moved the rate. Staffed intervention was a precondition, not an add-on.
- The audit export was originally pitched as "nice to have" — in production it became the biggest staff-time saver. Three person-weeks back per audit cycle is a measurable refinancing.
Could your business look like this in 4 weeks?
30-minute free discovery call. We review your setup and project realistic KPI ranges — honestly, even if less than in this case.
This case study is anonymised and based on typical project parameters of comparable Superchat implementations. It is not an individual guarantee of any specific results. Actual results depend heavily on the starting point, market segment and implementation discipline.