In short
Freelancers cost more time than they save when the business has nothing written down, because every brief has to be assembled verbally by the person who is already the constraint. Fix it before hiring anyone: document the one process they will run, define what an acceptable output looks like, and give them the context that is not in the task. Then treat them as a standing relationship rather than a series of one-off engagements, because the second and third jobs with the same person are where the return on that setup actually arrives.
Key takeaways
- The briefing cost lands on your most constrained person, which is why external help can reduce net capacity.
- Write the standard down before the first job. Otherwise the work comes back different and gets judged as wrong.
- Give context, not just the task. Externals cannot absorb it from the room, because there is no room.
- One freelancer used ten times beats ten used once. Setup cost is paid once and returns on every repeat.
- Plan the bench before the peak. A contractor found in a crisis is briefed in a crisis.
Why external help often costs time
The theory of bringing in a freelancer is straightforward: there is more work than capacity, so add capacity. The practice frequently disappoints, and the reason is where the additional work lands.
Bringing someone external in creates four costs, and every one of them falls on the person who knows the work best, who is almost always the person already at capacity.
- Briefing. Explaining the task and the context, from scratch, verbally, because nothing is written down.
- Availability. Answering questions during the work, on the freelancer's schedule rather than yours.
- Review. Checking the output, which requires the same expertise that would have done the work.
- Correction. Fixing what came back different, or explaining the difference and waiting for a second version.
In a business with no documentation, those four can consume more time than the task. The freelancer has done real work, and the constrained person has spent their week on briefing and review, so total output barely moves. This is the same dynamic that makes junior hires expensive before they are cheap, described in should you hire or fix the process first, compressed into a shorter timeframe.
Capacity added downstream of your constraint does not add capacity. It adds work to the constraint, in the form of briefing and review.
What has to exist before the first brief
Three artefacts, none of which takes long, and all of which are worth having whether or not you ever use a contractor.
| Artefact | What it prevents |
|---|---|
| A written procedure for the one process they will run | Rebuilding the brief verbally for every engagement, and for every new contractor |
| A definition of acceptable output: two or three attributes | Work coming back different and being experienced as wrong |
| A context pack: what the client cares about, what went wrong before, house conventions | The category of failure that looks like carelessness and is actually missing information |
The middle one is the one people skip, and it produces the most friction. A standard that has never been written down is a standard the contractor cannot hit except by luck, and the resulting judgement, that this person is not very good, is usually unfair and always expensive, because it triggers a search for a different contractor and the whole setup cost is paid again.
The method for producing the first artefact quickly is in how to write an SOP people actually follow: capture it live from whoever runs it now, then cut everything that is not a decision or a threshold.
What a brief must carry
A brief that only contains the task is the commonest cause of an unusable first delivery. Five elements, none of them long:
- The outcome, not the activity. What this is for and how you would know it worked.
- The standard. The two or three attributes, plus one example of previous work you were happy with and a sentence on why.
- The context that is not in the task. The client's sensitivity, the thing that went wrong last time, the convention everybody internally knows. Employees absorb this from the room. There is no room.
- The one failure that must not happen. Every piece of work has one. Naming it lets the contractor calibrate their caution rather than being cautious about everything.
- Who answers questions, and how fast. A single name and a realistic response time. Without it, contractors guess rather than ask, because chasing a client feels like admitting difficulty.
That last point is worth emphasising for external relationships specifically. A freelancer has a commercial incentive to appear self-sufficient, so the social cost of asking is higher than it is for an employee. If asking is not made explicitly easy and explicitly welcome, you will receive assumptions instead of questions, and you will find out which assumptions at delivery.
Treat the first job as calibration
The first engagement with any contractor is partly a test of them and largely a test of your briefing. Structure it accordingly.
- Make it small and real. Not a trial task invented for the purpose, which tests nothing about how your actual work behaves.
- Set a checkpoint at roughly a quarter of the way through. Early enough to correct direction cheaply. Reviewing only at delivery converts a briefing gap into a full rework.
- When the first version differs, ask which brief produced it. Most differences trace to something you did not say. Fix the brief, not the person.
- Write down what you had to explain. That list is next time's context pack, and it is the highest-value document this engagement will produce.
Engagement is where an external relationship starts paying. The first covers setup; the second and every one after it runs on documentation that already exists. Businesses that use a different freelancer each time pay the first-engagement cost permanently.
Build a bench, not a search
The most expensive way to use contractors is to find one in a crisis. Under time pressure you select from whoever is available, brief them badly because there is no time, and review their work badly for the same reason.
The alternative takes an afternoon a quarter:
- Identify the one or two roles you predictably need at peaks. Your capacity model tells you which, and when.
- Find people before you need them and run one small real job with each, at a calm moment.
- Keep the relationship warm. An occasional small piece of work keeps context alive and keeps you in their schedule.
- Keep the context pack current, so a peak engagement starts with a document rather than a call.
This also changes the economics of your capacity plan. Staffing to the average and flexing the peak only works if the flex is genuinely available and briefed; otherwise the plan quietly assumes a resource that does not exist.
What not to hand out
Three categories, and the boundary is not about difficulty.
- Work requiring judgement you have never articulated. If the rules exist only as accumulated instinct, an external cannot apply them and will be blamed for the gap. Write them down first, at which point the work becomes outsourceable.
- Anything reaching a client without your review. Not because contractors are careless, but because the accountability stays with you regardless of who produced it.
- Processes still changing week to week. An unstable process is a specification that moves after work has started, which is expensive with employees and worse with someone billing by the hour.
That third category is the same constraint that governs automation: fix and document before you hand off, whether the recipient is a tool or a person. The sequencing argument in automation maturity applies almost unchanged, because both are cases of moving work out of the heads of the people who currently hold it. The Mayim Ops assessment scores documentation and process clarity precisely because they determine whether adding hands adds output.
Frequently asked questions
Why do freelancers sometimes create more work than they save?
Because briefing, reviewing and correcting fall on whoever knows the work best, who is usually the person already at capacity. If nothing is documented, every brief is assembled from scratch verbally, and the time spent explaining plus the time spent correcting can exceed the time the task would have taken.
What should you give a freelancer before they start?
The task, the standard that makes an output acceptable, the context that is not in the task, and the name of one person who answers questions. The context is the part most often omitted and the part that most reliably determines whether the first delivery is usable.
How do you brief a contractor effectively?
State the outcome rather than the activity, give one example of previous work you were happy with and say why, name the single failure that must not happen, and set a checkpoint early enough to correct course. A brief without an example forces the contractor to guess at a standard you have never articulated.
Is it better to use one freelancer repeatedly or several occasionally?
One repeatedly, in almost every case. The cost of an external relationship is front-loaded into context transfer, and that cost is paid again with every new person. The second and third jobs with the same contractor are where the setup investment is recovered.
What work should not be outsourced?
Anything requiring judgement calls the business has never articulated, work that carries your accountability to a client without you seeing the output, and processes that are still changing week to week. An unstable process handed to an external is a specification that changes after the work has started.