Why Indonesian upskilling is a scale problem before it's a content problem

Indonesia has 280 million people and a workforce that's young, mobile-first, and moving online faster than most planning decks assume. By 2023, more than 55% of executive and employee training in the country already ran online or blended. The regional learning management system market generated about $6 billion in 2024 and is growing past 20% a year. So the question for an HR director at a Jakarta conglomerate isn't "should we go digital." That ship sailed. The question is how to spend a finite budget across a workforce too large to train the old way, one classroom at a time.

That's the frame that matters. Indonesian upskilling is a logistics and economics problem first, a curriculum problem second. Get the delivery model and the co-funding right, and content is the easy part.

Related reading: How Philippine IT-BPM Firms Are Reskilling for GenAI in 2026 · Malaysia's Digital Talent Strategy and MDEC Incentives in 2026 · Corporate Training and AI Upskilling in Central & Eastern Europe 2026.

Kartu Prakerja: the co-funding lever most enterprises underuse

Kartu Prakerja, launched in 2020, is Indonesia's national skills card. It gives citizens over 18 subsidised access to skilling, reskilling, and upskilling courses through approved platform partners — Tokopedia Academy, Skill Academy, Pintaria, and others. It was built for individuals, which is why most corporate L&D teams file it under "not our problem."

That's a missed lever. The Microsoft Indonesia partnership with Kartu Prakerja engaged 84,505 learners, a proof point that public skilling rails can carry serious enterprise volume when a company plugs into them deliberately. Pair that with the Kominfo Digital Talent Scholarship, which runs free AI and data-science tracks for Indonesian professionals, and you have two state-funded pipelines feeding the exact skills your enterprise is paying full price to teach. A smart 2026 strategy routes entry-level and foundational skilling through these subsidised tracks and reserves the private training budget for role-specific, company-specific capability that no public course can deliver.

Build, buy, or co-fund — how to actually split it

Here's a split that holds up across most large Indonesian employers I've seen plan for 2026:

  • Foundational digital and AI literacy — route through Kartu Prakerja partners and the Digital Talent Scholarship. Subsidised, broad, good enough for baseline. Don't pay full price for what the state already funds.
  • Role-specific technical skills — buy from specialists. Bangkit Academy (Google-backed), Dicoding Indonesia, and Pertama Partners for customised in-house AI workshops all sell genuine depth. This is where vendor quality earns its premium.
  • Company-specific capability — build internally. Your product, your systems, your compliance rules. Nobody sells this off the shelf, and outsourcing it is how institutional knowledge leaks.

The mistake conglomerates make is paying premium vendor rates for tier-one literacy that Kartu Prakerja covers, then under-investing in the company-specific layer that's actually their competitive moat. Flip that.

The conglomerate advantage nobody talks about

Indonesia's economy runs on large diversified groups, and that structure is a hidden upskilling asset. A worker reskilled in data analytics inside a group that spans banking, telco, retail, and logistics can be redeployed across business units without leaving the parent company. The retention math is different from a standalone firm's. Astra, Sinarmas, Lippo, GoTo — groups at that scale can treat upskilling as internal talent mobility rather than a cost center, moving a reskilled person from a shrinking function to a growing one instead of hiring outside and laying off inside.

If you run L&D inside a group like that and you're still planning per-subsidiary, you're leaving the biggest efficiency on the table. The unit of planning should be the group, not the company.

What the 2026 data actually says about urgency

The pressure is real and dated. Indonesia's skills challenge isn't new, but its scale in 2026 forces a recalibration: too many workers, too fast a shift, too little time to do it classroom by classroom. The government keeps expanding large-scale training precisely because the private sector alone can't move fast enough. GovInsider documented the state pushing workforce productivity through mass training programs for exactly this reason.

For an enterprise, the takeaway is timing. The subsidised rails are funded and running now. The companies that wire into them in 2026 get years of state-funded skilling before the programs inevitably tighten eligibility or shift focus. The ones that wait will pay full freight later.

A measurement habit that survives contact with reality

Most Indonesian upskilling programs measure completion. Completion is vanity. The metric that matters is redeployment: how many reskilled workers actually moved into a higher-value role within six months. A bank I know in Jakarta ran 2,000 people through a data-analytics track and celebrated a 91% completion rate, then discovered fewer than 200 had changed what they did day to day. The training worked. The deployment didn't. They'd built a certificate factory, not a capability pipeline.

Track redeployment from day one. If your LMS can't tell you who moved roles after training, it's measuring the wrong thing, and so are you.

The first 90 days

Start by auditing what Kartu Prakerja and the Digital Talent Scholarship already cover, then strike that content from your paid plan. Whatever's left is your real training spend, and it's smaller than you think. Point it at role-specific and company-specific capability, set redeployment as the success metric, and if you sit inside a conglomerate, plan at the group level. The companies that win Indonesia's talent decade won't be the ones that trained the most people. They'll be the ones that moved the most people into work that pays more and lasts longer.