How We Got Trapped — Government SW Pricing, Hancom, and 30 Years of SI Outsourcing

How public SW pricing, Hancom/groupware NIH, and 30 years of SI outsourcing produced Korea's build lock-in.

How Korea’s Software Market Got Trapped in Build

In Part 0 we saw the contradiction: Korea ranks #51 in software spending but #1 in R&D. This article retraces how that came to be — through 30 years of history.

Build lock-in is neither accidental nor innate, nor is it about conservatism. It’s the cumulative result of three forces: public software pricing policy from the 1990s, the in-house development preference symbolized by Hancom and groupware, and three decades of IT outsourcing. Three axes that reinforced one another and hardened over time.

Axis 1 — The Yoke of Public SW Pricing

Korea’s software market was driven from the 1990s by public procurement. RFPs from government ministries, public agencies, and financial institutions grew the market.

The problem lies in the structure of pricing. Public SW project cost calculation hardened around a headcount-input model.

“To build this system, N developers × M months” is the starting point of pricing
The unit of measurement is how many people, for how long, not how much value
License-based, subscription-based, or outcome-based pricing struggles to fit

Inside this model, packaged software and SaaS feel awkward to price. SaaS doesn’t fit “5 people for 6 months.” As a result, public commissioning standardized around build (outsourced headcount input), and that standard influenced private commissioning.

Public commissioning pricing made build the standard, and the private sector followed.

Hancom vs M365 — A Classic NIH Case

In the same period, the symbol of in-house SW development preference in Korea became Hancom Office. Globally, Microsoft Office became the de facto standard. Korea, centered on the public sector, kept Hancom Office.

This wasn’t simple NIH (Not Invented Here). Inside the public commissioning pricing model, “foreign license fees” were hard to justify, and adoption of in-house-developed Hancom Office was politically defensible. In other words, public SW pricing structure and in-house development preference reinforced each other.

The same pattern repeated in groupware, ERP, and CRM. While SaaS standardized globally, Korea kept commissioning custom systems through SI. This is the first mechanism that turned Korea into “a market that has SI build packages” rather than “a market that buys packages.”

Axis 2 — Three Decades of IT Outsourcing

In the late 1990s, through the IMF crisis, Korean conglomerates and financial institutions rapidly outsourced their IT. They spun out internal IT into subsidiaries or transferred them to SI affiliates, slimming HQ IT down to “planning and operations.”

Two outcomes followed:

Sparse in-house engineering teams: HQs lost the capacity to build directly.
Outsized SI industry growth: Samsung SDS, LG CNS, SK C&C — the SI Big 3 — took on group-wide IT and grew into massive businesses.

Era	Model	Outcome
Early 1990s	In-house IT departments	Some self-development, some foreign adoption
Late 1990s	Outsourcing · spinoffs	In-house engineering weakens, SI subsidiaries grow
2000s	Era of the SI Big 3	Group IT depends on SI, foreign SaaS struggles to enter
2010s	Cloud arrives	Some SaaS attempts, but SI model still dominates
2020s	AI arrives	sLLM, in-house build attempts collide with SI outsourcing inertia

This 30-year accumulation is the structure Korea faces today. In-house engineers can’t build directly because there aren’t enough of them, and commissioning SI is build, but a different mechanism from in-house build of the SaaS or AI era.

Self-Reinforcement of “Custom Build” Culture

As SI outsourcing became standard, one cultural norm hardened: “Our company’s processes are special, so standard SaaS won’t fit.” This is precisely the difference Younglimwon Soft Lab discovered in Japan (detailed cases in Part 2). Japanese mid-market companies adopt 1,000+ screens of SaaS ERP without changing a line of code. Korean companies demand large-scale customization for the same SaaS.

This isn’t because companies are conservative. It’s because 30 years of working under a build model leaves people unfamiliar with how to “work to a standard.” Organizations habituated to outsourced commissioning reflexively respond to SaaS with “make it fit our process.”

Axis 3 — Custom Build Practices for Groupware and ERP

Conglomerates and the public sector preferred to commission groupware and ERP through SI rather than buy SaaS, sticking to building their own systems. Three reasons:

Pricing model: SaaS subscription feels awkward inside public commissioning pricing.
Security perception: Direct deployment to internal networks feels “safer” (combined with network separation policy).
Headcount-input evaluation: Receiving N people × M months from SI produces visible deliverables.

A byproduct of this practice is regulation like CSAP and network separation. When a market standardizes self-build, that standard hardens into security regulation. Foreign SaaS struggles to enter on top of those regulations. The entry barriers weren’t created as separate policy; they emerged naturally from market practice.

In this sense, Korea’s build lock-in isn’t a simple “policy failure.” Commissioning practice creates security regulation, and security regulation in turn protects commissioning practice — a self-reinforcing loop is at work.

Summary — A Self-Reinforcing Structure Built Over 30 Years

%%{init: {'look': 'handDrawn', 'theme': 'neutral'}}%%
graph TD
    A[Public SW Pricing<br/>= Headcount Input] --> B[Build Model Standardized]
    C[IT Outsourcing] --> D[Sparse In-house Engineering]
    D --> E[SI Industry Outsized Growth]
    B --> F[Self-build · NIH Culture]
    F --> G[CSAP · Network Separation Regulations]
    E --> H[Foreign SaaS Entry Barrier]
    G --> H
    F --> I[Customization Demand]
    I --> J[Korean SaaS Can't Grow at Home]
    H --> J
    style A fill:#fef3c7,stroke:#d4a373
    style C fill:#fef3c7,stroke:#d4a373
    style J fill:#ffd6a5,stroke:#c98a5a

The structure isn’t any single point but multiple variables reinforcing each other, accumulating. Pricing model → SI outsourcing → sparse in-house engineering → self-build culture → regulation formation → SaaS entry blocked → Korean SaaS companies blocked from growing.

Era	Variable	Cumulative Effect
1990s	Headcount-input pricing	Build model standardized
Late 1990s	IT outsourcing	Sparse in-house engineering
2000s	SI Big 3 outsized	Group IT dependent on SI
2010s	Self-build → network separation · CSAP	Foreign SaaS entry blocked
2020s	Customization demand culture	Korean SaaS companies can’t grow at home

Next — The Cost of Being Trapped

The next installment looks at the cost this trap creates today. Foreign SaaS can’t enter, and Korean SaaS can’t grow at home — two faces of the same mechanism. We see this in data through three Korean SaaS companies detouring through Japan: LINE WORKS, Younglimwon, RSUPPORT.

Sources

Younglimwon Soft Lab — SystemEver SaaS ERP (Korean) — SaaS policy, adoption cost
Public SW Project Cost Guide — NIPA Korea Information & Communications Industry Promotion Agency (Korean) — Headcount-input pricing model
Korea Software Industry Association (KOSA) Statistics (Korean) — General SW industry data

Part 1’s 30-year macro narrative synthesizes industry-recognized history rather than relying on single primary sources. Specific stats are verified through primary sources in Part 2’s cases.

Series Index: Korea Build Trap