Most offshore engagements that fail were visibly going to fail during evaluation, if anyone had asked the right questions. The information was available. It just was not requested, because evaluation tends to focus on portfolio and price, and problems live elsewhere.
This checklist is deliberately one you can run against us. Publishing the criteria we work to is more useful than another article explaining that communication matters.
Key Takeaways
- Portfolio and price are the least predictive signals. Both are easy to present well.
- Ask who specifically will do the work. Attrition between pitch and delivery is a common failure.
- Ownership and exit are contract questions, not trust questions. Settle them before signing.
- A short paid pilot beats every reference call. Real work on a real problem is the only reliable evidence.
- Vague answers are the signal. What a vendor cannot answer precisely tells you more than what they can.
Team and Continuity
1. Who specifically will work on this, and can you meet them before signing?
2. Are they employees or subcontractors? Subcontracting is not disqualifying, but undisclosed subcontracting is.
3. What is the team’s tenure at the company? High churn means your context leaves repeatedly.
4. Who is the named technical lead, and do you speak to them directly or through an account manager?
5. What happens if that lead leaves mid-engagement?
6. How many other projects will these people be on concurrently?
Working Model
7. How many hours of daily overlap with your team are committed — and is it in the contract or an intention?
8. What are the working hours in their timezone, and does the shift arrangement apply to the whole team or just the lead?
9. What is the escalation path when something is urgent outside those hours?
10. Which tools do they work in, and will you have direct access to the backlog and the repository?
11. How is progress reported, and how often?
Technical Verification
12. Can you see actual code, not just screenshots of finished products?
13. Who wrote the portfolio piece you found most impressive, and are they still employed there?
14. What is their test coverage practice, and can they show it on a real repository?
15. Is code review mandatory before merge, and how is that enforced technically?
16. How do they handle dependency updates and security advisories?
17. Will they do a short paid pilot on a real, bounded problem? A refusal is informative.
Commercial and Legal
18. Who owns the repositories, cloud accounts and domains during the engagement — you or them?
19. What exactly transfers at the end, and how long does handover take?
20. What is the notice period, and what does it cost to exit early?
21. For EU personal data: is there a data processing agreement, are Standard Contractual Clauses in place, and has a transfer impact assessment been done where required?
22. What is their position on certifications such as SOC 2 or ISO 27001 — held, in progress, or not pursued? Any of those is an acceptable answer; evasion is not.
Judgement
23. Have they told you something you did not want to hear? A vendor who has never pushed back during evaluation will not push back during delivery.
24. Ask what kind of project they are a poor fit for. An inability to answer means either poor self-knowledge or a willingness to take work they should decline.
25. Ask about a project that went badly and what changed as a result. The specificity of that answer is the single most informative question on this list.
How to Weight the Answers
Not all twenty-five carry equal weight. If you are short of time, the four that predict most:
- Can you meet the actual team? (1, 4)
- Is the overlap contractual? (7)
- Do you own the infrastructure from day one? (18)
- Will they say no to something? (23, 24)
A vendor who answers those four precisely will usually answer the rest well. A vendor who answers them vaguely will not improve after signing.
The Total-Cost Question
Rate comparison is the most misleading part of offshore evaluation. A rate 60% lower does not make delivery 60% cheaper once you account for rework, coordination overhead, and your own senior engineers spending mornings unblocking people.
That does not make offshore a bad decision — it makes rate a bad way to evaluate it. Offshore is genuinely cheaper when scope is stable and the vendor is strong, and most expensive on exploratory product work where requirements move weekly. Our comparison of nearshore, offshore and onshore models works through the total-cost model, and the regional rate guide covers what the raw numbers actually look like.
How NSDBytes Answers These
We publish this checklist because we would rather be evaluated on it than on a portfolio page.
On the four that matter most: you meet the engineers before signing, not just a lead. Overlap hours are written into the contract rather than left to goodwill. Your repositories, cloud accounts and domains are registered to you from day one — we work inside your environment and there is nothing to transfer at the end because nothing was ever in our name. And we will tell you when we are the wrong fit, including when the answer is a local vendor.
On certifications, the honest answer: we hold neither SOC 2 Type II nor ISO 27001. If that is a hard gate in your procurement process, you should know now rather than three weeks in. What we do instead is set the terms out in the contract, where they are enforceable.
Our engagement models describe how teams are structured commercially, and hire a professional covers the dedicated-team arrangement specifically.
Final Thoughts
Run this list. Run it against us, and against everyone else you are considering.
The questions that produce hesitation are the ones worth pursuing. A good partner will answer twenty-five uncomfortable questions without difficulty, because the answers are just how they already work — and the ones who cannot are telling you something useful, for free, before you have signed anything.
If you want to work through it with us directly, get in touch.



