How to Evaluate a Vietnamese Offshore Vendor — 5 Axes the Sales Deck Never Shows
- Problem: Offshore vendors' sales decks are all rates, headcount, and big client logos — with not a line about what actually decides whether your project succeeds.
- Solution: Score vendors on 5 axes — BrSE maturity, whether quality processes actually exist, accountability structure, team retention, and pilot design ability — and screen with 8 red flags.
- Result: You avoid the "cheap turned expensive" trap, structurally preventing rework and project-on-fire costs that dwarf a 20–30% rate difference.
When choosing a Vietnamese offshore development company, don't decide by rate card and portfolio. What separates success from failure is five factors that never appear in the sales deck: BrSE maturity, whether quality processes actually exist, the accountability structure, team retention, and pilot design ability.
TL;DR (Executive Summary)
- Problem: Vendor proposals contain only rates, headcount, and big client logos — you can't read who owns outcomes or how quality is assured.
- Solution: Score on 5 axes (BrSE maturity / real quality processes / accountability structure / retention / pilot design) and cut early with 8 red flags.
- Result: You escape the "cheap turned expensive" trap — rework and firefighting costs are an order of magnitude larger than a 20–30% rate gap.
I've spent 14 years as a Japan–Vietnam BrSE and Delivery Manager, standing on both the client side and the vendor side, evaluating vendors and rescuing projects. This article is the exact framework I use in the field.
Why a sales deck can't tell you anything about a vendor
Vietnamese offshore vendors' proposals almost always revolve around three things: "low rates", "engineer headcount", "big-name client logos". None of the three correlates meaningfully with whether your project will succeed.
70% of outsourcing failures are caused by broken accountability, not technical skill, and most quality problems are standards-transmission problems. So what you should evaluate isn't "what they can build" — it's "how information travels, and who sits inside the accountability structure".
The 5-axis vendor scorecard
| Axis | How to verify | Passing bar |
|---|---|---|
| ① BrSE maturity | Interview the actual assigned BrSE | Can recount really pushing back on a spec |
| ② Real quality processes | Ask to see the actual DoD / review criteria | Real artifacts from past projects appear |
| ③ Accountability structure | Ask "who owns it if this fails?" | Instant answer with a name and role |
| ④ Team retention | Ask each proposed member's tenure | Core members ≥ 2 years |
| ⑤ Pilot design ability | Request a small-start proposal | Pilot comes back with clear validation goals |
Axis ①: BrSE maturity — "interpreter" or "owner"?
The line "1 BrSE" in a quote tells you nothing. Always interview the actual BrSE who will be assigned and ask: "When did you last say No to a client request?" No concrete example means a translator-type BrSE. I've written the detailed classification in the 4-level BrSE maturity model.
Axis ②: Real quality processes — don't accept "we have them", say "show me"
"We do rigorous quality management" carries zero information. Ask to see the actual Definition of Done, review criteria sheet, and test reports from past projects (confidential parts masked is fine). A company with real processes produces them in minutes. A company that can't — their process exists only in the sales slides.
Axis ③: Accountability structure — "when the project burns, whose phone rings?"
Ask directly: "If this project fails, who in your company owns that?" Answers like "the whole team" or "the company takes responsibility" are red flags. Only a company that can instantly name a specific individual (a Single Point of Accountability) will function when the project catches fire.
Axis ④: Team retention — will the proposal's lineup still be there in 3 months?
Vietnam's IT job-hopping cycle is fast — the ace from the proposal being gone before development starts is not rare. Ask each proposed member's tenure and the handover rules when someone leaves (documentation, handover period). A vendor that can describe anti-key-person-dependency as a mechanism is one you can trust.
Axis ⑤: Pilot design ability — can they start small, intelligently?
Don't sign a large contract immediately; ask the vendor to propose a 1–2 month small pilot. What to look at isn't the pilot's price but the quality of its design: what is this pilot meant to validate (communication accuracy? quality process? domain understanding?). A vendor that defines validation goals clearly will run the main project as hypothesis-driven, too.
The 8 red flags (one hit = dig deeper)
- Instantly answers "can do" to every request (zero pushback)
- Abnormally cheap quote (rework cost isn't priced in)
- Reluctant to let you interview the assigned BrSE directly
- "Needs time to prepare" when asked to show real process artifacts
- Answers "the whole team" when asked who owns failure
- The proposal's lineup differs from the contract's lineup
- Can't recount a single past failure (zero failures doesn't exist)
- Suggests adding more engineers before the contract is signed (a man-month business tell)
Frequently asked questions
Are large vendors or small vendors safer?
Size is not a safety indicator. Large vendors have stable structures, but your project risks becoming "one of many" staffed by juniors. Small vendors may commit their aces, but retention risk rises. Either way, scoring all 5 axes — especially "interview the actual assignees" — absorbs the noise that size introduces.
What if we have no one technical enough to evaluate vendors?
A client side with no one who can evaluate technically is the most disadvantaged structure in vendor selection. In that case, bringing in a third party independent of the vendor (a client-side BrSE or someone with delivery experience) to review proposals, interview the BrSE, and assess the pilot removes the information asymmetry.
If you're at the vendor-selection stage feeling "the proposal looks great, but I have no basis to trust it", feel free to share your case. I typically respond to cases that fit my expertise and current availability.