Hiring Office Help or Sending the Work Out: What Changes

Both routes work. What separates a good outcome from a disappointing one is almost never which was chosen — it is whether the owner understood what they were taking on, because each route replaces the work with a different kind of work rather than removing it.

Start from the functions, not from the route

Owners usually approach this backwards. They decide that the answer is a person, or that the answer is sending things out, and then try to work out what that person or that arrangement should do.

That order produces the two most common failures. Deciding on a person first produces a hire with a vague brief who ends up absorbing whatever is loudest, which is not the same as relieving the owner. Deciding to send work out first produces a specification nobody can write, because the work was never named precisely enough to describe to somebody who cannot see it.

Starting from a written list of functions makes most of the decision evaporate. Some of what is on that list obviously needs someone present and answerable in the moment. Some of it is bounded, repeatable and better performed by whoever does nothing else. Once the list exists, the two routes stop competing and start dividing.

What attaches when somebody joins the payroll

The part owners most often meet after the decision rather than before it is the standing obligation.

Taking someone onto the payroll makes the company an employer for federal tax purposes, with a recurring set of duties — withholding, deposits, returns — that run on their own calendar regardless of whether the month was good. The IRS sets these out in Publication 15, its employer tax guide, which exists precisely because the obligations are ongoing and detailed enough to need a book.

None of that is an argument against employing anybody. Plenty of companies should. It is an argument against treating the decision as purely a question of whether the wages are affordable, because the wages are the part that was already visible. The rest is a permanent administrative function the company did not previously have, and somebody has to run it.

The management load is the cost nobody budgets

The second unbudgeted cost is attention, and it lands in the worst possible period.

Somebody new is not useful on arrival. They need decisions, context, and correction, and every one of those has to come from the person who understands the work — which is the overloaded owner. For the first stretch the arrangement consumes more of that person’s capacity than it returns.

This is not a criticism of anybody’s competence. It is arithmetic, and it is why hiring is a strong answer to next quarter and a weak answer to this week. An owner who brings somebody in during the month they are drowning has added a training obligation to a month that was already the problem, and both the owner and the new person will conclude something has gone wrong when nothing has.

The management does not stop after the first few weeks either. It becomes smaller and permanent: a person needs someone to be accountable to, decisions escalated to, and time from. That is an ongoing job with the owner’s name on it.

What sending work out actually costs

The mirror-image cost on the other route is specification, and it is just as real and just as easy to leave out of the reckoning.

An outside desk does what was described. It cannot infer a preference, and it cannot know about the customer who has always been handled differently or the supplier who needs chasing on a Thursday. Anything that lives only in the owner’s head is simply absent from the result, and its absence reads as poor performance when it is actually missing information.

The work of getting that out of the owner’s head is front-loaded, unfamiliar and boring, and it is the single best predictor of whether the arrangement will be judged a success. Companies that do it well spend an uncomfortable amount of time at the start describing things they have never had to describe. Companies that skip it get an accurate rendering of an incomplete brief.

The compensating property is that the specification is durable. It does not leave, it does not need managing, and it holds when the volume changes. Which of the two front-loaded costs — training a person or describing the work — is the better investment depends entirely on how long the arrangement is expected to last.

Nothing about either route is instant

An owner reaching this decision is usually past the point of comfort, and both routes take longer than that state of mind wants.

Recruiting, choosing and settling somebody takes months. Specifying and handing over a function takes weeks. Neither is available on the timescale of the week that finally prompted the decision, which is exactly why the decision keeps getting deferred: it is only ever considered at moments when there is no capacity to act on it.

The way out of that loop is to accept that the decision belongs to a quiet period and to make it there, in advance of needing it. That is unsatisfying advice and it is the only version that works.

The route nobody counts: moving it inside the team you have

There is a third possibility that gets skipped because it does not feel like a solution, and for a certain size of company it is the correct one.

Most roofing businesses already contain somebody who is doing a job that is partly administrative and partly something else — a production coordinator who also handles suppliers, a lead estimator who also answers the phone in the afternoon. Those roles accumulated the same way the owner’s did, by absorbing whatever needed doing, and they are frequently carrying work that would be better placed elsewhere in the same team.

Redistributing is genuinely cheaper than either alternative. There is no recruitment, no obligation attaching that was not already there, and the people involved already understand the business. What it costs is the thing owners are worst at spending: a decision that somebody will stop doing something they currently do, which usually means telling a capable person that a task they have quietly owned for years is moving.

It fails when it is done implicitly. Asking somebody to “help out with” a function while they retain everything else is not redistribution — it is the same work with a second person occasionally touching it, which adds coordination without removing load. The version that works names the function, moves it whole, and takes something else off the person receiving it.

The honest limit is capacity. If the team is already at its ceiling this route redistributes strain rather than relieving it, and doing that repeatedly is how a company loses the people it most needs. But it is worth testing before either of the more expensive answers, because it is the only one available this month.

The mixed answer is usually the right one

Framing this as a single choice is what makes it hard. Very few companies of this size are best served by moving everything one way.

The functions that suit somebody employed tend to share properties: they need presence, they involve judgement about people, or they require being interruptible. The functions that suit being run elsewhere share different ones: they are bounded, they are repeatable, they are damaged by interruption, and they benefit from being done by whoever does nothing else all week. The service pages set out the second group in detail, function by function, which is a more useful way to read them than as a menu.

An owner who divides the list this way usually finds the answer is both, in a specific ratio, and that the ratio is obvious once the functions are written down.

What to do before deciding anything

Write the list. Every recurring administrative thing the company does, with a name, and next to each one a note about whether it needs a person present.

Take the two items that cost the most attention and decide those two only. Not the whole office — two functions, chosen because they hurt most. That produces a real decision with a real outcome, which is worth more than a complete plan nobody executes.

The rest of this cluster is about why the work collected on one person to begin with, and what has to be written down before any of it can move. This guide is about the moment after that, when the owner has decided something has to change and is choosing what kind of change to make.