How to Choose the Right Partner for the Job

Choosing the right partner for a job comes down to five questions: What's your failure rate? Who exactly will do the work? What happens if it goes wrong? Can I speak to a client from last year? What's not included in the price? Ask those before you ask about cost.

At peptest, founded in 2019 on Marszałkowska 42, we've watched clients make the same hiring mistakes for six years running. This post is what we'd tell a friend. It's also what we wish more people asked us.

What questions should I ask before hiring a specialist?

Most people ask first about price. That's the wrong first question. Price tells you almost nothing about outcomes, and cheap work redone twice costs more than good work done once. A 2023 review published on PubMed found that project rework driven by poor supplier selection accounts for a significant share of total project cost overruns [1].

Ask these instead:

• What percentage of your projects finish on the original timeline and budget, and can you show me the last ten?
• Who on your team will actually do my work, and how many years have they been doing it?
• What's your policy when something goes wrong after delivery, in plain writing?
• Can I contact two clients you served twelve months ago, not last week?
• What's explicitly excluded from the quote you're about to send me?
• How do you handle scope changes mid-project, and what triggers a new invoice?
• What would make you turn down my job?

That last one matters most. A partner who's never turned down work is a partner without standards. When we said no to a client in March 2024 because their timeline was unrealistic, they thanked us six months later. They'd hired someone else. It didn't go well.

How do I compare quotes from different providers fairly?

Quotes look similar on paper and behave nothing alike in practice. You need a comparison structure, not a gut feeling. Here's the frame we hand our own clients when they're vetting us against two competitors.

| Criterion | Solo freelancer | Small studio | Large firm |

|---|---|---|---|

| Typical cost | 2 000–5 000 PLN | 6 000–15 000 PLN | 20 000+ PLN |

| Response time | Hours to days | Same business day | 2–5 business days |

| Best for | Small, defined jobs | Ongoing work needing continuity | Compliance-heavy or multi-team projects |

| When to avoid | Anything mission-critical | Very small one-off tasks | Under tight budgets |

Read the table sideways, not top to bottom. The cheapest column is only cheap if the job fits that column. A freelancer at 3 000 PLN who ghosts you halfway through costs more than a studio at 9 000 PLN who finishes on time. We covered this pattern in more depth in our notes on how we approach client work at peptest.

Why do credentials matter less than you think?

Here's the "actually" moment. Credentials tell you someone passed a test. They don't tell you if they can do your specific job in your specific situation. We've hired certified professionals who couldn't deliver, and we've watched self-taught practitioners run circles around them.

What matters more: recent, relevant, referenceable work. Recent means within the last eighteen months. Relevant means the same category of problem, not just the same industry. Referenceable means you can pick up the phone and talk to the person who paid them.

Last October a client came to us convinced she needed a total rebuild because a certified vendor had told her so. Two hours of review later, we'd found the actual problem was a single misconfigured setting. Cost to fix: 400 PLN. Cost the other vendor had quoted: 18 000 PLN. Credentials aren't proof of judgment.

For public-facing standards you can verify independently, the European Commission publishes procurement guidance that's useful even for private contracts [2]. Google's own vendor evaluation criteria for search partners are also public and worth reading before any digital hire [3].

Try this: before your next meeting with a potential partner, write down the three outcomes you actually need. Not features. Outcomes. Then ask them to explain, in their own words, how they'd get you there. If they can't, walk. If you want a longer read on how we structure client conversations, see our welcome post at peptest.

FAQ

How many providers should I get quotes from?

Three is the sweet spot. One quote gives you no reference. Two forces a binary choice. Four or more creates decision paralysis and slows you down by weeks. Pick three, brief them identically, and compare their questions back to you as carefully as you compare their prices.

Should I always pick the most experienced partner?

No. Experience matters, but fit matters more. A twenty-year veteran who's bored by your project will deliver worse work than a five-year specialist who finds it interesting. Ask directly whether your job excites them. Watch the answer, not just listen to it.

What's a red flag I should never ignore?

Vague answers about who does the work. If a partner can't name the specific person who'll handle your project, you're buying a promise, not a service. Subcontracting isn't automatically bad. Hidden subcontracting is.

How long should the vetting process take?

For jobs under 10 000 PLN, one week is plenty. For larger commitments, give yourself two to three weeks including reference calls. Rushing this step is the single most common regret we hear from clients who came to us after a bad first hire.