Lab-Type vs Contract: A 2026 Guide to Choosing Your Vietnam Offshore Development Model

"We don't have enough engineers in-house — but we have no idea which development model to use when outsourcing." This is the first wall many Japanese companies hit when they try to outsource system development. When considering Vietnam offshore development in particular, there is one choice you cannot avoid: the choice between the contract-based model and the lab-type model. This article lays out the essential differences and how to choose in light of 2026 trends, from the perspective of the person placing the order.

Why "choosing the model" is now a management issue

According to Japan's Ministry of Economy, Trade and Industry (METI), the country's IT talent shortage is projected to reach up to roughly 790,000 people by 2030 (the current shortfall is around 170,000). IPA surveys also show that over 60% of companies report a severe shortage of DX talent. Competition in the hiring market is intensifying, and securing development resources through domestic hiring alone is becoming unrealistic.

Against this backdrop, offshore development — the continuous use of overseas development resources — is increasingly positioned not as a cost-cutting tactic but as a talent strategy for business continuity. And the first factor that determines success or failure is which contract model you choose.

The two offshore development models

Contract-based development

The contract model assumes delivery of a completed deliverable. The client fixes the specifications, the vendor designs, builds, and tests according to those specs, and delivers by the deadline. The client does not direct the development process directly; instead, they receive regular progress reports from a bridge SE or project manager.

In short, it is a model where you can "hand it over completely, with delivery date and quality guaranteed by contract." It suits projects with clear specs that complete in one shot (replacing part of an existing system, building a new business system with fixed requirements). The downside: it is weak against mid-development spec changes, which tend to trigger additional fees and re-estimates.

Lab-type development

The lab model is a contract to secure a dedicated development team for a fixed period. You retain engineers on a monthly basis and continuously direct that team. You pay based on capacity (per person-month) rather than per deliverable.

Its greatest strength is flexibility. Even with fluid requirements, you can reshuffle priorities each sprint. Because domain knowledge accumulates within the team, productivity rises the longer you work together. It pairs excellently with MVP validation, products that need continuous post-release improvement, and agile development. On the other hand, the client must have the capacity to manage the team (articulating specs, setting priorities, communicating). Neglect this and you fall into the trap of "having a dedicated team but getting no results."

Contract vs lab: side by side

The 2026 cost reality: "because it's cheap" is no longer a reason

As for cost, Vietnam offshore rates for a programmer hover around 300,000 to 400,000 yen per person-month as a rough guide (varying with skill and whether a bridge SE is included). If domestic development cost is indexed at "100," some estimates put Vietnam offshore at around "60 to 70" — a 30 to 40% cost reduction.

But rate is not the only reason Vietnam is chosen. The country has over 550,000 software engineers and a steadily growing ICT market. With a diligent workforce and strong affinity with the Japanese market, Vietnam's position has shifted in recent years from "a cheap vendor" to "a strategic technology partner."

One caveat: urban labor costs in Vietnam are rising year by year, and the once-overwhelming price gap is narrowing. 2026 may well mark the end of choosing purely on "cheapness."

Which should you choose: five decision criteria

• Are the specs fixed? Fixed: contract. Fluid: lab-type.
• One-off or continuous? One-time: contract. Ongoing post-release fixes: lab-type.
• Can you spare internal management effort? If not, contract. If you can assign a dedicated owner, lab-type.
• Do you want to accumulate know-how continuously? If long-term knowledge build-up matters, lab-type.
• How do you budget? Want a fixed project total: contract. Want stable monthly costs: lab-type.

The 2026 trend: generative AI x offshore

As of 2026, a development style that uses generative AI to cut the work hours themselves is becoming mainstream. By adopting coding-assistant AI and AI agents, the same headcount can produce dramatically more output.

This means future offshore partner selection must add "productivity via AI utilization" as a comparison axis, not just the simple "per-person-month rate." A low rate with low productivity ends up expensive; a slightly higher rate with a team that masters AI can be cheaper in total. Because the lab model accumulates know-how within the team, it pairs well with turning AI tools into internalized organizational capability.

The key to success: what to nail before the model

Whichever model you choose, what ultimately decides success is communication design. Hand over specs and context while they are still vague, and even an excellent team will not deliver what you expected. Bring in the Japanese practice of "Ho-Ren-So" (report, contact, consult) as a shared language, and use the bridge SE to catch misunderstandings early — this unglamorous operational design is where offshore development is truly won or lost.

Conclusion

The contract model suits companies that want to "hand over a one-off project with fixed specs, with delivery guaranteed." The lab model suits companies that want to "stay resilient to change and grow development capability over the long term." 2026 is a turning point where the structural challenge of IT talent shortage and the productivity revolution of generative AI advance at once. Reading your own project characteristics and choosing the optimal model not only on cost but on "productivity" and "continuity" will be the key to successful system development going forward.