How to Choose an IT Contractor: What Actually Predicts a Good Engagement
The most common mistake in hiring IT contractors isn't choosing the wrong person. It's choosing for the wrong reasons.
Rate, portfolio aesthetics, and how well someone presents in an intro call are the things most clients evaluate. None of them reliably predict whether a project will be delivered on time, on scope, and without a support dependency that outlasts the contract.
Here's what does.
Before you talk to anyone: get your brief right
Contractors fail more often because of unclear client requirements than contractor-side incompetence. A PMI Pulse of the Profession study found that 29% of projects fail due to poor requirements gathering at the outset. That's not a contractor problem — it's a scoping problem that the client owns.
Before reaching out to anyone, write a one-page project brief that answers:
- What problem are we solving? (Not what are we building — what problem does it solve?)
- What does done look like? (Specific, testable outcomes)
- What is explicitly out of scope?
- Who on our side is the decision-maker?
- What's the feedback process and turnaround expectation?
Contractors who receive a clear brief give more accurate estimates. Contractors who receive a vague brief give optimistic estimates that create problems later.
What to evaluate — and what to ignore
| Weak signal | Why it misleads | Stronger signal |
|---|---|---|
| Portfolio aesthetics | Shows taste, not delivery reliability | Portfolio relevance — did they solve similar problems? |
| Intro call confidence | Good presenters aren't always good builders | How they handle an ambiguous requirement in real time |
| Rate (lowest) | Lowest rate often means most optimistic scope assumptions | Rate relative to the specificity of their estimate |
| Years of experience | Time in field doesn't predict communication quality | Reference check: did they hit timelines and flag problems early? |
| Certifications | Demonstrates platform knowledge | Demonstrated outcomes in similar context |
The reference check questions that actually work
Most reference checks produce polite non-answers because clients ask the wrong questions. These produce useful information:
"Did they deliver on the agreed timeline? If not, what happened and how did they handle it?" The second half matters more than the first. Problems arise in every project. How a contractor handles problems is what distinguishes them.
"What would you do differently if you worked with them again?" People will not volunteer criticism unprompted. This question creates permission to give it. Listen for anything about communication lag, scope surprises, or handoff quality.
"How much time did your team spend managing them versus working alongside them?" High-maintenance contractors are expensive even at low rates.
The proposal stage
Two things to look for in a proposal beyond price:
What are they assuming? Two proposals at different price points are often solving slightly different versions of your problem. Find the assumptions that differ and evaluate which is more realistic.
Can they produce a milestone plan on request? A contractor who has done this type of work before should be able to produce a basic timeline showing deliverables, dependencies, and review points within 24 hours of a request. If they can't, that's an answer about how the project will be managed.
Contract essentials
| Element | Why it matters |
|---|---|
| Written Scope of Work | Defines what's in and what's a change order — protects both sides |
| Milestone-based payment | Aligns incentives: they get paid when you get deliverables |
| IP assignment clause | Confirms code and asset ownership transfers to you at delivery |
| Revision and feedback process | Defines what counts as a revision vs. a scope change |
| Termination terms | What happens if either party needs to exit cleanly |
The step most clients skip: the paid pilot
A two-week paid test on a small, bounded task tells you more than any interview or portfolio review.
How do they communicate when they hit a problem? How do they interpret requirements that are slightly unclear? Do they deliver what they committed to, on the timeline they proposed? Do they flag when something is taking longer than expected, or do they go quiet?
The cost of a two-week pilot is small relative to the cost of discovering six weeks into a large engagement that the working relationship doesn't function. Every client who has skipped this step and regretted it will tell you the same thing.
On using platforms like Upwork
Platform reputation metrics are imperfect but not meaningless. Expert Vetted status on Upwork covers the top 1% of contractors based on verified client outcomes — it requires a real track record, not just a good profile.
Job Success Score above 90% sustained across multiple diverse clients is a meaningful signal. A single five-star review from a two-week engagement is not.
Read the reviews for the specific things that appear in multiple separate reviews. If five different clients mention "communicated proactively when there was a delay" — that's a real pattern. If five different clients mention "delivered on time" — that's the most reliable predictor you'll find before working together.
We've been on Upwork for years with a 100% Job Success Score and Expert Vetted status. We mention this not as a credential but because the reviews describe how we actually work — specifically, how we handle scope ambiguity and what happens when something unexpected comes up.
The engagements that go well almost always start with a clear brief and a well-defined first milestone.
If you're evaluating contractors right now and want to understand how we approach scoping before anything is signed, that conversation is free.
agency.pizza →
