Choosing a nearshore dedicated development team is not a logo exercise. The risk is not that someone can write code in a demo. The risk is that six months later you still do not own the product knowledge, the delivery rhythm, or a partner you can hold to a contract.
That is why vetting has to happen before the first sprint — on the company, the way they work, and the people who will actually sit in your tools. D-Factor runs a five-stage process on every partner in the network. Most applicants do not pass. You only see teams on the other side of that filter.
This article is both how we do it and a checklist you can reuse if you are evaluating any software development partner in Poland and EU nearshore markets.
Why “we have 200 developers” is not vetting
Volume is not a quality signal. A bench of CVs, a Clutch page, and a slide about “senior engineers” tell you almost nothing about:
- Whether the company will still exist and stand behind a replacement
- Whether code review, hiring, and attrition are real processes or slogans
- Whether the people on your Slack have been through an engineer-led screen
- Whether you can enforce IP, billing, and EU-side contracts when something goes wrong
Unvetted freelancers and anonymous marketplace profiles fail this test by design. Staff augmentation with a thin vendor layer often fails it quietly: you get a person, not a team you can inspect. For one to three specialists on a short horizon that can still be the right buy — see Outstaffing / Staff Augmentation. For a unit on your roadmap for six months or more, you need a team-level filter.
The five stages below are that filter.
Stage 1 — Reputation review
We start with the company, not a candidate PDF.
- Public track record and references — not only selected case-study quotes
- Conversations with previous and current clients where we can get them
- Consistency across sources: delivery claims vs what clients actually say about communication, stability, and ownership
What you should copy: ask for two client conversations in a similar engagement shape (team size, duration, domain). “We cannot share names” is sometimes real; “we cannot share anyone” is a stop sign.
Reputation is a gate, not a score. A polished website does not compensate for a pattern of churned clients or unexplained team swaps.
Stage 2 — Process audit
A dedicated team lives or dies on how work actually moves — hiring bar, review, incidents, knowledge sharing — not on a certified wall of logos.
In the process audit we look at how the partner:
- Hires and levels engineers (who owns the technical bar)
- Runs code review, QA, and release discipline
- Handles absences, holidays, and replacement without resetting your board
- Documents enough that product knowledge does not live in one person’s laptop
- Collaborates with a client who keeps product and architecture governance
Certificates can support this. They cannot replace it. If the partner cannot describe a normal week on a client account, they are selling capacity, not a team.
This stage is also where outstaffing vs dedicated team becomes visible. Individual placement with you absorbing all rituals is outstaffing. A composed unit with account management, stability, and a shared working culture is a dedicated development team.
Stage 3 — Technical interviews
People still have to pass as engineers.
We review skillset, code quality, relevant domain and stack, then run interviews that cover:
- Technical depth at the seniority you are buying — not title inflation
- How they reason about trade-offs, not only syntax
- Communication: can they work with your PM, EM, and existing in-house staff
Recruiters schedule. Engineers own pass/fail. The same principle we use in IT recruiting applies here: a partner who cannot discuss the stack with your tech lead is forwarding CVs, not vetting a team.
For a dedicated team you are not scoring one hero. You are checking that the unit can hold a bar together — seniors who can lead, mids who can own slices, no “remote discount” on quality.
Stage 4 — Test project
A conversation is not delivery. A real test project shows how the team estimates, splits work, reviews each other, and ships something inspectable.
We are not looking for a weekend puzzle. We are looking for:
- Working habits under a small, time-boxed brief
- Whether “senior” means judgment or just years on a CV
- How they handle incomplete requirements — because your backlog will look like that too
Buyers evaluating a partner without a network should do a paid, scoped spike (days, not months) before a six-month retainer. Unpaid “homework” that looks like free production work is a different smell; keep the spike small and explicit.
Stage 5 — Partnership review
Only after the first four stages do we accept a team onto the verified shortlist. Partnership is the commercial and operating fit:
- EU-side contracting (D-Factor is headquartered in Poland), IP assignment, invoicing
- Who you call when delivery slips — a named counterpart, not a ticket queue
- Replacement and continuity terms
- Whether they accept that you own the product: roadmap, priorities, and architecture stay with you
A partner who wants to disappear into a black-box “we will handle everything” is a different product (classic outsourcing). Dedicated team, in our model, means you keep direction; we compose and govern the capacity. More on that split: Dedicated Team vs. Outstaffing.
What buyers should ask on the first call
Use these even if you never speak to D-Factor:
- Who technically screens — engineers or sales?
- Do you vet the company and the people, or only send CVs?
- Can I speak to a client with a similar team shape?
- What does replacement look like in week twelve?
- Where does product ownership sit on day one — and on day ninety?
- Is this a dedicated unit or staff augmentation with a team label?
- What fails your process? (If the answer is “almost nobody,” there is no process.)
If you need the engagement model itself, not the filter: three or more engineers, six months or longer, one counterparty → Estimate My Team. One to three specialists, shorter horizon → Outstaffing.
After they pass: onboarding is still on you
Vetting reduces the chance you hired a theatre company. It does not replace kickoff.
Access, operating model, and a codebase brief in week one still decide the next six months. We wrote that playbook separately: How to Onboard a Dedicated Dev Team in Two Weeks. If outstaffing has already started to cost you coordination and attrition, the model switch is here: When to Stop Outstaffing and Move to a Dedicated Team.
FAQ
Is five-stage vetting only for dedicated teams?
The partner network is five-stage vetted for dedicated teams and for specialists we place through outstaffing. The engagement still differs: unit + stability vs individual capacity.
Can we interview the engineers ourselves?
Yes. If you want to approve candidates yourself, assembly takes longer (typically into the 4–8 week range). Standard composition from the vetted pool is faster because the filter already ran.
Do you use freelancers?
No. Specialists come from vetted partner companies — employment, accountability, and a replacement path sit behind each person.
Where are the teams?
Poland and a partner network across Europe, structured for timezone overlap with the UK, Western Europe, and Nordics. Contracts under EU law. We do not sell an anonymous offshore pool.
Decide the filter, then the model
- Inspect the partner — reputation, process, people, a spike of real work, commercial terms
- Choose the model — dedicated team vs outstaffing vs permanent hire
- Keep product control — backlog, architecture, and success metrics stay yours
See the five stages on About us, then talk to us about a composed team: Dedicated Development Team.