In short
- Decide what you are buying before you decide who to buy it from.
- Pick the engagement model from the shape of the work, not from habit.
- Judge people on questions asked and work shown, not on how fast they reply.
- Make the first commitment small enough that being wrong is cheap.
Most bad freelance hires are decided before anybody applies. The brief was vague, so the quotes were guesses; the work was never split up, so the first payment carried all the risk; the shortlist was built on response speed rather than on anything the person had done. This guide covers the decisions that matter, in the order you actually face them.
Work out what you are buying
There is a real difference between hiring an outcome and hiring time. An outcome is a thing you can look at and accept: a landing page, a migrated database, twelve product descriptions. Time is capacity you direct: someone who spends ten hours a week on whatever the sprint needs.
If you can describe the finished thing in a sentence, you are buying an outcome, and you should be able to agree a price. If the work will change week to week, or you cannot name the deliverable yet, you are buying time, and pretending otherwise produces a fixed price that one of you will resent.
- Name the deliverable, or admit that you cannot yet.
- Write down what finished looks like, in terms you could check.
- Decide who makes the final call on quality, and make sure that person is available.
- Gather what the freelancer will need on day one: access, brand files, content, examples.
Choose how the work is structured
Four shapes cover almost any job, and the right one is usually obvious once the work has been described honestly.
| Shape | Use it when | How money moves |
|---|---|---|
| Fixed-price | The deliverable is known and can be split into stages | You fund each milestone before that stage starts, and release it when you approve the work |
| Hourly | Scope will move, or you are directing the work week to week | You fund a week at a time up to an agreed hour limit, approve the diary, and unused funding goes back |
| Dedicated Hire | You want ongoing monthly capacity from one person | A monthly engagement rather than a project |
| Drop-In | The work is standard and someone already sells it at a set price | You buy the package as listed, with its stated delivery time and revisions |
The trade-off between the first two is the one worth thinking about, and fixed-price versus hourly goes through it properly.
Write a brief people can quote against
A brief is not a specification. It needs to carry enough for a stranger to estimate cost and time without a meeting: what exists now, what should exist afterwards, what is already decided, and what you will supply.
Weak
Need a website. Send your price.
Better
We need a six-page marketing website for a SaaS product. Figma designs are ready. The developer should build it in Next.js, connect the contact form to our CRM and make sure the pages work well on mobile.
The second one can be priced in a few minutes, because it settles the three things that move a quote the most: how much there is, whether design is done, and what it has to connect to. The first will get a range so wide it tells you nothing, or a low number that grows once the work starts. How to write a freelance job post has the full structure.
Read the replies properly
Proposals punish skimming. The useful signals are quiet ones: a question that shows somebody read the brief, an example resembling your problem rather than their best-looking work, an estimate split into parts instead of a single number.
- A good question beats a good pitch. Every brief has a gap. Someone who finds yours is thinking about the work.
- Relevant beats impressive. A modest project in your situation is worth more than a spectacular one in a different one.
- Look at how they price. A number split into parts can be discussed; a single number can only be accepted or refused.
- Watch the writing. You will be reading this person's updates for weeks.
Reply speed is the signal to ignore hardest: it measures who happened to be at a screen. How to review freelance proposals covers building the shortlist.
Start smaller than the whole job
The most reliable way to test a working relationship is to run a small piece of it. Make the first milestone a real, useful chunk with an output you can judge: a first page, a schema, three of the twelve articles. On a fixed-price contract only that milestone is funded when work starts, so an unsuccessful trial costs the price of the trial.
Mistakes that cost the most
- Hiring on price alone. The cheapest quote is often the one that understood the least.
- Leaving quality undefined. If nobody wrote down what good looks like, review turns into taste.
- Disappearing after kickoff. Freelancers stall waiting for access, feedback and decisions. Name a person who answers.
- Paying outside the platform. Money sent off-platform is not in escrow and cannot be recovered through a dispute.
- Scope creep by message. New work is a new milestone, not a favour.
How much should I budget?
It depends on scope, seniority and how much is already decided, so ask for quotes against a written brief rather than a range in your head. What you can control is how much you commit at once: a small first milestone gives you real information before the rest of the budget is on the line.
How do I know the freelancer will deliver?
You do not, in advance, which is why the structure matters. On Open Lance the money is funded into escrow before work starts and released when you approve it, so neither side has to trust the other with the whole project. Reviews on this platform are tied to completed contracts, and verified stats from other platforms are shown on the profile.
Can I hire the same person again?
Yes, and it is the cheapest hiring you will do. Repeat work needs no shortlist and usually no brief. Start a new contract with them directly.
What if the work is not what I asked for?
Request revisions first: that is what the review window is for. If you cannot agree, either side can open a dispute while the milestone is still funded, and the money stays frozen until it is resolved. Once a milestone has been released it can no longer be disputed.