Use contractors for variable demand, specialist skills needed occasionally, and testing whether a role is genuinely permanent - not as a cheaper permanent employee, because per hour they are more expensive and the saving is in commitment rather than cost. Brief them the way you would brief a client: written scope, defined deliverable, named approver, agreed rate and turnaround. The two failures that account for most contractor problems are a vague brief and an unclear position on who owns the client relationship, and both are fixable in the first conversation.
Almost every agency uses contractors, and most manage them worse than they manage clients.
The brief is verbal. The scope is "help us with the build." The rate was agreed in a message. Nobody said whether they speak to the client directly. And then when it goes wrong - a deliverable that missed the point, a bill higher than expected, an awkward moment in a client call - the conclusion is that contractors are unreliable.
They usually are not. They were briefed badly, which is a solvable problem and entirely yours.
This guide covers when contractors are the right answer, how to find and evaluate them, the brief that prevents most problems, rates and payment, the client-facing question, and how to build a bench you can rely on.
When a contractor is right
Variable demand. Work that spikes and subsides. Hiring permanently against a peak leaves you carrying the cost through the trough.
Specialist skills used occasionally. Motion design four times a year, a particular integration, accessibility auditing. Genuine expertise you cannot justify full-time and should not attempt in-house.
Testing a role. Six months of contract work tells you whether the need is permanent far more cheaply than a hire that turns out to be wrong.
Covering absence. Parental leave, long illness, a departure mid-project.
Buying time to hire properly. Recruiting well takes months. A contractor keeps delivery moving without forcing a rushed permanent decision.
When a contractor is wrong
As a cheaper employee. They are not cheaper per hour - typically 1.3 to 2 times the equivalent employee rate. What you are buying is flexibility, and paying for it. Anyone reaching for contractors to reduce cost has the model backwards.
For work needing deep institutional knowledge. Account leadership and client-facing project management usually belong in-house, because they depend on knowing your systems, your other clients and your commercial position.
As a permanent arrangement in disguise. A contractor working full-time for you for two years, on your systems, to your direction, is a permanent employee in most jurisdictions' eyes - with real tax and employment-status consequences. If the relationship has become permanent, make it permanent.
When you have no capacity to brief. The most common practical failure. Contractors need more explicit direction than employees, not less, because they lack the context employees absorb passively. If nobody has two hours to brief properly, adding a contractor makes this week worse.
Finding and evaluating
Referrals from other agencies are the best source by a distance. An agency that has used someone can tell you how they handle feedback and deadlines - the two things that actually matter and that no portfolio reveals.
Former colleagues and past employees. They already know how you work.
Communities and networks in your discipline, over generic marketplaces, where filtering is expensive and signal is low.
Evaluate with a small paid piece of real work. Not a portfolio review, not an unpaid test - a small, paid, representative task. You learn how they interpret a brief, how they respond to feedback, and whether they communicate when stuck. Those three predict everything and are invisible in a portfolio.
Ask about their other commitments. Someone with three concurrent full-time engagements is not available in the way you need, and finding that out in week three of a deadline is expensive.
The brief
Where most contractor problems originate. Brief a contractor at least as well as you would brief yourself, and preferably the way a good client briefs you.
Six elements, written:
The deliverable, specifically. Not "help with the site" - "six page templates designed to the existing design system, delivered as source files, with responsive behaviour defined at three breakpoints."
The context. Who the client is, what the project is for, what has already been decided and why. Contractors produce work that misses the point mainly when they were never told the point.
The constraints. Brand guidelines, technical stack, accessibility requirements, things that have been tried and rejected.
Who approves, and how many rounds. A named person and a stated number of revision rounds, exactly as you would agree with a client. Unbounded revisions are as damaging to a contractor relationship as they are to yours.
The rate, the estimate, and what happens if it overruns. Agreed in writing before work starts. State explicitly whether they should stop and check at the estimate or continue - this single sentence prevents most billing disputes.
Deadlines, including yours. When you will provide feedback. Contractors are frequently blocked by agencies, and then blamed for the resulting delay.
If that list looks like a statement of work, it is, and for the same reasons.
Rates and payment
Expect to pay 1.3-2x the equivalent employee hourly cost. They carry their own overhead, tax, insurance, holiday, pension, downtime between engagements and business development. A contractor rate that looks equivalent to a salary is not.
Day rates are usually cleaner than hourly for defined work. Fewer arguments about fifteen minutes, and easier to plan around.
Agree the estimate and the overrun rule up front. Covered above and worth repeating, because it is the single most common source of contractor friction.
Pay promptly. The highest-return practice in this entire guide. Contractors have no HR function to chase for them and cash flow matters enormously to a single person. An agency that pays within seven days gets first refusal on availability, better rates, and goodwill during a crisis. An agency that pays at 60 days is at the bottom of everyone's priority list, whatever they say about the relationship.
That is also the honest version of a cash flow trade-off: paying contractors quickly costs you working capital and buys you reliability. For most agencies it is worth it.
The client-facing question
Decide it explicitly, before the first client interaction.
Three workable models:
Invisible. The contractor never contacts the client; you brief, review and present. Maximum control, maximum overhead on you, and it wastes the contractor's ability to ask a direct question.
Visible as part of the team. They join calls under your banner, introduced as part of the team. Common, efficient, and it needs your client agreement to permit subcontracting - check what your contract says before assuming.
Named as a specialist partner. Introduced explicitly as an external expert. Suits genuine specialists and can strengthen your positioning, since bringing in named expertise reads as diligence rather than as a gap.
All three are fine. What is not fine is leaving it undecided, so the contractor guesses on a call and either overstays their remit or refuses to answer a simple question and looks evasive.
Two rules regardless of model: be clear about who owns the relationship - always you - and never let a contractor negotiate scope or price with your client.
Contracts and the practical essentials
Not legal advice, and four things belong in every arrangement.
IP assignment. The client is buying work that you are selling. If your contractor has not assigned rights to you, you cannot pass them on. This is the most commonly missed clause and the one most likely to cause a real problem.
Confidentiality, covering the client's information as well as yours.
Payment terms, stated, including what happens on late payment.
Non-solicitation, if you want it - and be realistic. A blanket ban on ever working with a client they met through you is often unenforceable and always resented. A time-limited, narrow clause is more likely to hold and less likely to sour the relationship.
Also check your own client contracts permit subcontracting. Some do not, and discovering that after the fact is genuinely awkward.
The economics, honestly
Agencies frequently reach for contractors as a cost measure and then find margin worse rather than better. Understanding why prevents the mistake.
The hourly cost is higher, deliberately. A contractor at £60 an hour against an employee whose fully-loaded cost is £35 an hour is not a saving. You are paying roughly a 70% premium for the right to stop at any time, and that premium is entirely rational on their side - they carry their own downtime, holiday, pension, insurance and business development.
The saving is in the commitment, not the rate. A contractor used for eight weeks costs 70% more per hour and 0% for the other forty-four weeks. Against an employee who would be 55% utilised, the contractor is cheaper overall. Against one who would be 85% utilised, they are considerably more expensive.
The break-even is roughly your target utilization. If a role would be busy above your utilization target, hire. Below it, contract. That single comparison answers the question more reliably than instinct, and it uses numbers you already have from utilization rate.
Budget for the briefing time. A contractor needs more explicit direction than an employee because they lack absorbed context. Two hours of briefing on a twenty-hour engagement is a real 10% overhead that never appears in the rate comparison and should.
Watch the margin on contractor-delivered work specifically. Calculate project margin separately for projects delivered by contractors. If it is consistently thinner, either your markup is too low or the briefing overhead is larger than you assumed - and both are fixable once visible.
Integrating them without friction
Two things determine whether a contractor works smoothly alongside your team.
Give them the same context your employees have. The temptation is to share only what is strictly necessary, which produces someone working blind. Share the client background, the commercial shape, the constraints and the history. A contractor who knows the project is fixed-fee treats scope differently from one who does not.
Include them in the rituals that affect their work. The kickoff if they are there from the start, the standup for the days they are working, the review that covers their deliverable. Excluding them and then briefing them separately costs more time than including them.
What not to do is treat them as staff in the ways that matter legally - setting fixed hours, requiring attendance at unrelated meetings, or directing how rather than what. Beyond the employment-status risk, it is also the fastest way to lose good contractors, who chose this arrangement deliberately.
Quality and review
Contractors need the same quality process as employees, and they frequently do not get it, because their work arrives as a finished deliverable rather than developing visibly in front of you.
Review contractor work against the brief, not against taste. The brief is the contract. If the work matches it and you dislike the result, the brief was wrong - and that is genuinely useful information about your own briefing.
Review early, not at the end. A checkpoint at 25% costs an hour and prevents a week of work heading the wrong way. This matters more with contractors than with employees, because you cannot see progress ambiently.
Give feedback in one consolidated round. Drip-fed comments over three days are expensive for someone billing by the hour and produce a worse result than a single considered set. The same discipline you would want from a client.
Be specific about revision rounds. Two rounds included, further rounds billed - agreed in the brief. Unbounded revisions damage a contractor relationship exactly as they damage yours with a client, and for the same reason.
When it is not working
Occasionally an engagement goes wrong. Handling it cleanly protects both the project and the relationship.
Raise it immediately and specifically. "The last two deliverables have missed the brief in the same way - can we get on a call?" Waiting until the end and then declining to use them again teaches nobody anything and leaves them confused about what happened.
Check the brief first. In a meaningful proportion of cases the contractor delivered exactly what was asked and the ask was unclear. That is your problem to own, and owning it visibly makes the correction land much better.
Stop early if it is genuinely wrong. Paying for work done and ending the engagement is cheaper than continuing and hoping. Say so directly and without blame - fit varies, and a contractor who is wrong for this project may be right for the next one.
Pay them properly on the way out. Including for work in progress. An agency that disputes a contractor's final invoice acquires a reputation in a small community very quickly, and it will cost you access to good people for years.
The handover when they leave
Contractors leave more often than employees, by design, and each departure carries knowledge out with it.
Ask for the same things you would give a client at project handoff: source files in editable form, a short note on decisions made and why, and anything a successor would need to know. Fifteen minutes of their time, billed, saves days later.
Build it into the engagement rather than requesting it at the end. A line in the brief - "a short handover note is included in the scope" - makes it expected and paid rather than an awkward favour asked after the final invoice.
Rates and markup
The commercial question agencies handle most inconsistently: what do you charge the client for contractor time?
Three approaches, all defensible:
Cost-plus markup. Contractor costs £60, you bill £90. Simple, and the markup covers your briefing time, review, project management and risk. 40-60% is typical and it is not profiteering - you are carrying the sourcing, the quality risk and the client relationship.
Your standard rate regardless. The client pays what they would pay for your own team, and the margin varies by who delivers. Cleanest from the client's perspective and it avoids any suggestion that they are buying a cheaper product.
Transparent pass-through plus a management fee. Used mainly for specialists the client has effectively asked for by name. Less common in agency work and appropriate where the contractor's identity is part of what is being bought.
What to avoid: billing contractor time at a rate that leaves you no margin, on the grounds that you did not do the work. You did the sourcing, the briefing, the review and the client management, and all four are real. An agency that passes contractor time through at cost is subsidising the engagement.
And be careful about disclosure. Some client contracts require you to declare subcontracting; some clients simply prefer to know. Check before assuming, because being discovered rather than telling them is a much worse conversation than the one you avoided.
Building a bench
The agencies that use contractors well have a small group they return to, and it does not happen by accident.
Keep them warm between engagements. A short message every couple of months. Availability goes to the people who stayed in touch.
Give notice where you can. "We are likely to need you in six weeks" is worth a great deal to someone planning their own capacity, and it costs you nothing.
Feed back honestly, both ways. Contractors rarely get told how work landed. Two minutes of specific feedback makes the next engagement better and is unusual enough to be memorable.
Include them in the retrospective where they did substantial work. They see things your team cannot, precisely because they are outside it.
Refer work you cannot take. The fastest way to become someone's preferred client.
Three or four reliable people, kept warm, is a meaningful competitive advantage - it means you can say yes to work you would otherwise turn down, and say it the same day.
The legal and status question
Not legal advice, and the risk is real enough to be worth naming plainly.
Most jurisdictions distinguish between a genuine contractor and someone who is an employee in all but name, and the tests are broadly similar: who controls how the work is done, whether the person can send a substitute, whether they carry business risk, how integrated they are into the organisation, and whether the arrangement is exclusive and open-ended.
The pattern that creates exposure is a contractor working full-time for one agency for a long period, on the agency's systems, to the agency's direction, with fixed hours and no other clients. That is an employment relationship with a different invoice attached, and the consequences of it being reclassified fall on the agency.
Practical mitigations, none of which are exotic: contract for defined deliverables rather than open-ended time, avoid setting fixed working hours, do not require attendance at meetings unrelated to their work, allow substitution where practical, and review long-running arrangements annually.
The honest test: if a contractor has been with you full-time for eighteen months and you would be seriously disrupted if they left tomorrow, they are functionally part of the team. At that point the right answer is usually to offer them a permanent role, which resolves the exposure and is frequently what both sides want anyway.
Check the rules in your own jurisdiction, and check what your client contracts say about subcontracting while you are at it - some prohibit it outright, and discovering that after delivery is genuinely awkward.
The summary
Contractors fail for two reasons: a vague brief and an undecided position on client contact. Both are fixed before any work starts.
Brief them like a client, agree the estimate and the overrun rule in writing, decide the client-facing model explicitly, and pay quickly.
Do those four things and contractors stop being a risk you take under pressure and become the thing that lets you accept work your permanent capacity could not.