Key takeaways
- Judge the people who will do the work, not the people who sell it.
- Run a small paid pilot before signing a long engagement.
- Overlap in working hours matters more than the hourly rate.
- Clear ownership of quality, in writing, prevents most disputes.
You have three proposals from nearshore firms, all with polished decks and similar rates, and no reliable way to tell which team will actually deliver. Sales presentations are a poor predictor of delivery. The better approach is to design a short evaluation that exposes how each partner works when the pressure is real.
Ask about the team, not the company
Company credentials tell you little about the specific people who will sit on your project. Ask who exactly will staff the work, how long they have worked together, and who replaces them if someone leaves. Ask to meet the delivery lead and the engineers or analysts directly, and see whether the sales contact can answer detailed questions or has to relay them.
- Who is the day-to-day lead, and how many engagements do they run at once?
- What is the typical tenure of the people on the team?
- How is knowledge handed over when someone rotates off?
- Which parts of the work would you subcontract, and to whom?
Test working-hour overlap honestly
Nearshore value comes largely from shared working hours. A partner in a nearby time zone can join your stand-up, answer a blocking question the same morning and review work before your day ends. Ask for the exact overlap window, not a vague promise, and check that the people you are meeting are the ones who will keep those hours.
Also ask how communication is structured. A written weekly update, a shared board you can inspect at any time and a named escalation path are worth more than frequent, unstructured meetings.
Run a small paid pilot
The most reliable test is a bounded piece of real work, paid at normal rates. For example, a two to four week pilot with a defined deliverable, a written acceptance checklist and a review at the end. You learn how they estimate, how they handle ambiguity, how they report bad news and how clean their output is.
- Pick a real but low-risk task with a clear finish line.
- Write acceptance criteria before work starts.
- Ask for a plan and estimate, then compare with what happens.
- Review the code, models or documents yourself or with an independent reviewer.
- Decide in advance what result earns a larger engagement.
Contracts, security and ownership of quality
Read the contract for the boring parts. Who owns the work product, where is your data stored and processed, who can access your systems and how is access removed at the end? Ask for their security practices in writing and how they handle your credentials. A serious partner answers these quickly and without discomfort.
Quality also needs an owner. Agree on what done means, how defects are reported and how long the partner fixes their own defects at no extra charge. Ambiguity here is where most disputes start.
Red flags worth walking away from
- Refusal to let you speak with the actual delivery team before signing.
- Pressure to commit to a long term before any pilot.
- Estimates that never change, or that cannot be explained.
- Vague answers on subcontracting, data location or access control.
- No visible process for reporting risks early.
