Last updated: 2026-07-14
The bottleneck in public-sector training is not features. It is headcount — one or two administrators covering thousands of employees.
When one or two administrators manage several thousand employees, blind spots are guaranteed. Distributing authority to departmental sub-administrators moves assignment, chasing, and tallying to the people who already know the staff — at one public institution, about 70 sub-administrators each covered 30–50 people. At audit time, completion status is pulled as a one-click report. The honest caveat: nothing guarantees all 70 operate to the same standard. Capability varies, it adds work on top of their day job, and whether it is tied to performance review is what actually decides quality.
One more thing we publish rather than hide: in our sample, all four public institutions saw MAU fall from 80–95% during compliance season to single digits off-peak. The one exception had loaded 2,000+ courses before launch (/en/data-report). What follows draws on 9 organisations (4 public institutions, 1 retail affiliate, 2 IT/gaming, 1 digital media, 1 educational institution). For the vendor-neutral view, see our 8 criteria for choosing an enterprise LMS.
19 out of 20 sports public enterprise employees log in every month, unprompted. Learning as habit, not obligation.
Korea Rail Network Authority has sub-admins in each of its 71 departments independently planning and running their own training.
The Health & Welfare HR Development Institute remotely trains 10,000 health and welfare workers nationwide.
A major public enterprise unified on-the-job training for 5,000 staff across nationwide branches.
HQ retains overall oversight, while departmental admins upload training for their own teams directly. Korea Rail Network Authority has 71 departmental admins running distributed training, with HQ monitoring the full status at a glance.
71 department admins — distributed operationsWhen auditors request "show training completion status," instead of 30 Excel sheets, just open one dashboard. Filter by department, course, or period and extract data instantly. Non-completers list and completion trends — all on one screen.
real-time dashboardRequired compliance training is auto-assigned, with step-by-step notifications sent to non-completers. The Health & Welfare HR Institute trains 10,000 users on pension law, editing content directly via the built-in editor. When regulations change, admins update and redistribute immediately.
Compliance training for 10,000 usersA sports public corporation runs 2,000+ service content items and recorded MAU 95% at peak. Institutions with 500+ courses show meaningfully higher off-peak MAU. Offer only 6 compliance training types, and learners have little reason to return once they finish.
MAU 95% at peakKorea Rail Network Authority's 71 departmental admins each register and manage their own team's training. The bottleneck of one HQ person uploading everything is gone. Non-completer tracking is faster, making pre-deadline nudges possible.
71 admins — distributed operationsA major public corporation aggregates 5,000 users' completion history in real time. When audit requests come in, departmental training status and periodic completion rates export with one button. Job-specific content views — like surveying team manuals — are tracked alongside.
Real-time aggregation for 5,000 usersOperating data from 9 organisations (4 public institutions, 1 retail affiliate, 2 IT/gaming, 1 digital media, 1 educational institution).
Median about 2,000 learners (60–3,000), median ~300 courses.
A baseline you can hold a vendor's proposal against. The figures that held and the ones that collapsed are both here.
| Metric | Value | Sample & conditions | Source |
|---|---|---|---|
| MAU during compliance training | 80–95% | All 4 public institutions | Operating data report |
| MAU off-peak | single digits | All 4 public institutions. We publish this rather than hide it | Operating data report |
| Sub-administrator distribution | ~70 people | Public institution B · 30–50 employees each. No guarantee all 70 operate to one standard | Customer stories |
| The exception that held off-peak | MAU 95% at peak | Public institution A · 2,000+ courses loaded at launch — an investment most institutions cannot make | Operating data report |
| Content threshold | 500+ courses | Institutions above this held meaningfully higher off-peak MAU. Course composition may matter more than the number itself | Operating data report |
| Cascading decline (retail affiliate) | 90% → 30% → 3% → 1% | ~3,000 employees · successive months after compliance training ended | Operating data report |
| Regular live rhythm | MAU ~73%/month | Digital media company · weekly Wednesday live; one internal video contest lifted MAU ~30%p | Customer stories |
| Public-sector headcount (context) | ~1.3M + ~400K | Public institution staff plus public enterprise/quasi-government staff — third-party statistic | Ministry of the Interior and Safety, 2024 |
Sources: https://www.touchclass.com/en/casehome · https://www.touchclass.com/en/data-report
The eight questions that come up most often in public-sector evaluations.
The difference is the operating model, not the feature list. Public institutions carry a heavy statutory training load — occupational safety, information security, harassment prevention — where 100% completion is an audit item, while one or two administrators cover thousands of staff. So the questions that matter are whether authority can be distributed to departments and whether completion evidence can be produced on demand. Source: https://www.touchclass.com/en/employee-communication
Yes. Departmental sub-administrators can be given scoped authority to assign courses, chase non-completers, and view their own department's status. At one public institution, about 70 sub-administrators each covered 30 to 50 people. One caveat we state plainly: distributing authority does not guarantee consistent quality across all 70. Capability varies, it is extra work on top of their day job, and linking it to performance review is usually what makes it stick. Source: https://www.touchclass.com/en/casehome
Completion status by course, department, and individual can be exported as a report without manual Excel work. Because the data is captured continuously rather than assembled at audit time, the evidence is ready when a request arrives instead of triggering a scramble. Source: https://www.touchclass.com/en/employee-communication
Yes. When the HR org chart changes, training targets can follow automatically, so a transfer or promotion does not require re-assigning courses by hand. This matters most in institutions with regular rotation cycles, where a manual roster is out of date within weeks. Source: https://www.touchclass.com/en/integration
Engagement does collapse, and we will not pretend otherwise: all four public institutions in our sample fell from 80–95% MAU during compliance season to single digits off-peak. The one exception had loaded more than 2,000 courses before launch. A retail affiliate with 3,000 staff slid 90% to 30% to 3% to 1% over successive months. The lever is having everyday content ready before the compliance deadline passes — not a feature. Source: https://www.touchclass.com/en/data-report
Running a large institution with only a few administrators is precisely the design goal. Central administrators set policy and courses; departmental sub-administrators handle assignment and follow-up for their own people. At one public institution roughly 70 sub-administrators covered 30 to 50 staff each, which shortened the time to track down non-completers. Whether it works depends on whether those 70 are given the time to do it. Source: https://www.touchclass.com/en/casehome
Yes. Announcements, leadership messages, and a feed can run alongside courses, so employees are not asked to check yet another tool. A digital media company holds a live session every Wednesday and averages about 73% MAU; a single internal video contest lifted MAU by roughly 30 percentage points. The risk worth naming: if the person who owns that routine leaves, the routine usually goes with them. Source: https://www.touchclass.com/en/casehome
TouchClass holds ISMS-P and ISO/IEC 27001:2022, with AES-256 encryption, IP-based access control, two-factor authentication, and admin activity logs. Knowledge assets created through our AI features are not used as AI model training data. Ask any vendor for the certificate itself, its scope, and its expiry date rather than accepting a yes. Source: https://www.touchclass.com/en/security
한국어 
日本語 
English 
