Data-Driven Learnerships: Measuring What Really Matters

Overview

• Completion rates show who finished a learnership, not whether it produced a skilled hire. Absorption and 12-month retention are the numbers that prove return on investment.
• A 2025 Bureau for Economic Research review found that only 57% of learners in substantive SETA programmes complete, and that absorption rates range from 6.1% to 83%.
• Track skills progression, dropout timing, the manager experience and post-programme outcomes in one system, and collect the data over WhatsApp while learners are still engaged.
• Give executives a cohort dashboard that answers five questions, led by absorption rate.

Why completion rates are the wrong headline metric

Completion rates tell you who finished a learnership, not whether the programme produced an employee your business kept. That gap matters, because the national numbers are weak on both counts.

A 2025 review of the SETA system by the Bureau for Economic Research (BER) found that throughput for learnerships, internships and artisan programmes is only 57%. The same review reports that tracer studies put absorption rates anywhere between 6.1% and 83%.

That range is the real story. Two programmes with identical completion rates can deliver very different business outcomes, and completion data alone will never show you which one you are running.

Which learnership metrics actually matter?

Seven metrics give a complete view of a learnership, and each one points to a decision you can act on. The calculations below keep them consistent across cohorts.

Attendance and completion still matter for SETA reporting. Treat them as hygiene data, not as proof of impact.

Where do learners drop out, and what does it tell you?

The month a learner leaves usually points to the part of the programme that failed. Tracking exits by month turns a vague retention problem into a specific fix.

Many early exits come down to missing information rather than low motivation. Transparent learnerships that set clear expectations from day one lose fewer learners in the first month.

Why measure the manager experience?

Host managers decide whether a learnership works on the floor, so their experience is an early indicator of learner outcomes. An overloaded manager produces disengaged learners months before the dropout figures show it.

Track four things for every host manager:

• Coaching time: hours spent with each learner per month.
• Feedback quality: whether feedback is specific and given on time.
• Work-readiness rating: the manager's score for each learner at every rotation.
• Team alignment: how closely learner tasks match the team's actual goals.

Keep collection under 30 seconds. A single-question pulse sent to a manager's phone after a coaching session gets answered, while a long form at month end usually does not.

How do you collect learnership data without adding admin?

Keep all learnership data in one system and collect it through the channel learners already use, which for most South African learners is WhatsApp. That removes the two biggest sources of admin: reconciling records and chasing responses.

Fragmented data is the default. Training providers, HR, host managers and learners each keep separate records, so nobody sees the full picture. A central platform such as Neptune holds attendance, assessments, progress reports, coaching notes and chat history against each learner.

Conversational tools such as txthr run check-ins, attendance confirmations, pulse surveys and support requests over WhatsApp. Responses land in the learner's record without anyone retyping them, which is how learnership admin scales past a few hundred learners.

Continuous contact also solves the hardest measurement problem: tracing learners after they leave. In a 2019/20 HWSETA tracer study, 70% of sampled learnership completers never responded, mostly because the contact numbers on record no longer reached them. A channel learners use every week during the programme is far easier to keep open afterwards.

What post-programme outcomes prove learnership ROI?

Four outcomes prove a learnership's value, and all of them are measured after the certificate: absorption, performance, retention and career progression. We recommend the time points below so results can be compared across cohorts.

Absorption also carries direct scorecard value. Under the Amended B-BBEE Codes, an entity can earn 5 bonus points for absorbing black learners, interns and apprentices at the end of their programmes, against a 100% target. Absorption only counts when the learner secures a long-term employment contract, so your definition should match.

If you also place youth through the YES programme, track both in the same system so absorption is reported the same way across programmes.

What should a learnership dashboard show executives?

An executive dashboard should answer five questions on one screen, cohort by cohort, and refresh from live data rather than month-end spreadsheets.

1. What is our absorption rate for this cohort, and how does it compare with the last intake?
2. Where are we losing learners, and why?
3. Are we on track for our B-BBEE skills development targets?
4. Which business units produce the strongest 12-month outcomes?
5. What does each absorbed learner cost compared with an external hire?

For question five, calculate cost per absorbed learner as total programme cost (stipends, provider fees and manager time) minus SETA grants and tax allowances received, divided by learners absorbed. That single figure lets a board compare the learnership with its other hiring channels.

graylink's reporting and insights tools build these views from live programme data. For the wider picture, see our guides to HR report types and recruiting software ROI.

Our view: lead with absorption, not completion

If you report one learnership number to your board, make it absorption rate per cohort. It is the only metric that links the programme to your talent pipeline, your B-BBEE scorecard and your cost per hire at the same time.

Completion, attendance and engagement still matter, but as early warnings that explain the absorption figure. Consider two cohorts. One completes at 90% and absorbs 10% of its learners, which makes it an expensive training course, while the other completes at 70% and absorbs 60%, which makes it a working hiring channel.

Most programmes cannot make that distinction today, as the tracer data above shows. Tracking outcomes while learners are still in the programme, rather than surveying them years later, is how you close that gap.

Frequently asked questions

What is a good learnership absorption rate in South Africa?

There is no single national benchmark. The BER's 2025 review found tracer studies reporting absorption anywhere from 6.1% to 83%, so your own previous cohorts are the most useful comparison. For the full 5 B-BBEE bonus points, the target is 100% absorption into long-term employment.

How long should we track learners after a learnership ends?

Track learners for at least 12 months after they complete. Absorption at programme end shows whether you hired them, but retention and performance at 6 and 12 months show whether those hires were worth making.

Do we need to replace our HRIS to track learnership metrics?

No. A specialised layer such as Neptune can manage learnership recruitment, tracking and check-ins, then push finalised employee records into your core HRIS through standard integrations. Your existing system remains the record of employment.