Choosing and running a nearshore QA engagement

Geography sets a band. Almost everything that moves a quote within that band is about the shape of the work rather than about where it is done.

• Seniority mix. A team that needs someone able to argue with an architect costs differently from one executing an agreed plan, and the difference is larger than any regional discount.
• Ramp-up. Every engagement has a period where the supplier is learning your product and producing little. Short engagements pay for that period at a much worse ratio than long ones.
• Reporting depth. Documented, auditable testing costs more than the same testing done informally, and in a regulated sector that documentation is the deliverable rather than an overhead.
• Environment quality. If a supplier cannot get a stable test environment, they spend billable time fighting it. This is the single most common source of budget overrun and it is almost always on the client side.

The cost material is covered in more detail in nearshore QA outsourcing in Poland. What this guide adds is the part that comes after you have accepted the number.

Most bad engagements are a model mismatch rather than a supplier failure. Embedded people put on a project-shaped brief will wait to be told what to do; a project team dropped into a daily standup will optimise for the ticket rather than the outcome.

Central European Time overlaps the whole Western European working day and gives a usable morning against the eastern United States. That is the actual claim, and it is worth being precise about what it buys.

It buys same-day resolution of ambiguity. A tester who hits an unclear requirement at ten in the morning can have an answer by eleven and keep working. The same question asked into a twelve-hour gap costs a day, and a project with three of those a week loses more than the rate difference that motivated the arrangement.

What it does not buy is a substitute for written decisions. Teams that rely on overlap to have conversations, rather than to resolve things that were written down badly, discover the limitation the first time somebody takes leave.

A supplier established inside the European Union is inside the same data-protection regime as an EU client. That removes an entire category of question rather than answering it: there is no third-country transfer to assess, no adequacy decision to rely on, no supplementary measures to document.

It does not remove the processing agreement, the sub-processor disclosure, or the obligation to handle test data properly — those apply regardless of where the supplier sits, and are covered in the guide to the EU compliance surface. What it removes is the transfer question, which is the one that most often stalls a procurement review.

1. 01Who specifically will do the work, and can I speak to them before signing? An engagement sold by one team and delivered by another is the most common bait-and-switch in this industry.
2. 02What does your first week depend on from our side, and who absorbs the cost if it slips? Suppliers who have not thought this through will bill the ramp-up either way.
3. 03How will I know whether this is working after a month? If the answer is a report rather than a change in something I already measure, that is a warning.
4. 04What would make you tell us to stop? A supplier who cannot answer has no line, which means they will keep billing past the point of value.
5. 05What happens to the work if we end the engagement? A suite only your people can maintain is a dependency dressed as a deliverable.

The last question is the one that separates a supplier from a lock-in. Ask it early and watch how comfortable the answer is.

Three cases, stated plainly because we would rather not take the engagement than take it and fail.

• The work is genuinely exploratory and the definition changes weekly. Any supplier relationship adds a contract boundary, and boundaries are expensive when the thing being bounded keeps moving.
• The knowledge that matters is tacit and undocumented, held by two people who have worked together for years. Transferring that costs more than the engagement saves.
• The real problem is organisational — nobody owns quality, or releases are political. Adding an external team to that produces a well-tested product nobody ships.

If the question is really whether the function belongs inside the company at all, that is a different decision and we have written about it separately in in-house QA versus outsourced QA.