The question usually arrives sideways. Bookkeeping has gotten away from you, or the marketing you've been doing at 10pm isn't good enough anymore, or you've turned down work twice because nobody could deliver it. So you start pricing the fix, and within an hour you're staring at two numbers: an agency's monthly retainer on one side, a salary range on the other.
Then you pick the smaller number, and roughly half the time you regret it.
The comparison isn't wrong because the numbers are wrong. It's wrong because price is the last thing that should decide this, and the two options aren't the same purchase. Hiring buys you capacity that is always there whether you need it or not. Outsourcing buys you output on demand and nothing in between. Those are different products, and which one is right has very little to do with which one is cheaper this quarter.
The Rate Comparison Is Rigged in Both Directions
Set the two side by side honestly and the first thing you notice is that neither headline number means what it looks like.
The outsourced rate looks high because it is a rate. It has to cover the provider's own overhead, their downtime, their tools, and their profit. Per hour, an outside specialist will almost always cost more than an employee doing the same task. That is the number owners fixate on, and it makes outsourcing feel like a luxury.
The salary looks low because it is only a salary. The real number includes payroll taxes, benefits, workers' comp, equipment, software seats, the recruiting cost to fill the seat, and the hours you or a manager spend directing the work. Depending on your state and what you offer, fully loaded cost runs meaningfully above the salary line — commonly a quarter to a third more. That's before you count the weeks the role is underused.
So the honest comparison isn't rate against salary. It's annual fully loaded cost of an employee against annual outsourced spend at your realistic volume. Run that and the gap usually narrows to the point where cost stops being the deciding factor at all — which is exactly what you want, because cost was never the right decider.
Four Questions That Actually Decide It
Once the money is roughly a wash, the decision comes down to the shape of the work. Four questions get you most of the way there.
1. Is the volume steady or lumpy?
An employee is a fixed cost. That's the whole trade. If you can honestly say this work will fill a real week, every week, for the next year, a fixed cost is the cheaper structure and the better one — the person gets faster, learns your business, and stops needing instructions. If the work comes in bursts, hiring converts a variable cost into a fixed one at exactly the wrong time. You'll pay for the quiet months, and the quiet months are when cash is already tight.
Be strict with yourself here. "I could keep them busy" is not the same as "this role has enough of its own work." Roles invented to keep someone busy become the roles you can't justify a year later.
2. How close is it to what customers pay you for?
The closer a function sits to the thing customers actually buy, the stronger the case for owning it. Your product, your service delivery, the relationships that decide whether someone renews — those are the source of your margin, and handing them to a vendor means renting your own advantage.
The further out you go, the weaker the case. Payroll processing, IT support, tax filing, cleaning, bookkeeping: standardized across every business in your industry, and a specialist doing it a hundred times a month will do it better than someone you train once.
3. Can you write down what good looks like?
This is the one owners skip, and it causes more failed outsourcing than price ever does. An outside partner executes a specification. If the specification lives only in your head as taste — "I'll know it when I see it" — you will spend more time rejecting work than you would have spent doing it.
Work you can define is outsourceable. Work that requires judgment you can't yet articulate has to stay close to you until you can articulate it. Sometimes the right answer isn't outsource or hire — it's write the process down first, then decide.
4. Could you actually supervise the hire?
Hiring into a function you don't understand is how owners end up paying a full salary for output they can't evaluate. If you couldn't tell a strong hire from a weak one in this discipline after six weeks, you are not ready to hire for it. You are ready to buy it from someone accountable for the result, and to learn the function on their dime.
Outsource what you can specify. Hire what you can supervise. Keep what you're paid for.
When the Answer Is Clearly Outsource
Some situations don't need a framework. Outsource when the work is genuinely specialized and genuinely occasional — legal, tax, a technical build. Outsource when you need the capability this month and hiring would take a quarter you don't have. Outsource when the volume is unpredictable or seasonal and you'd be paying for idle capacity. Outsource when the skill is one you'd never be able to interview for competently.
And outsource when you're buying a decision rather than a pair of hands. You don't hire a full-time employee to help you evaluate one acquisition or restructure your pricing once. You buy expertise for the duration of the question.
When the Answer Is Clearly Hire
Hire when the work is continuous, the volume is proven, and you keep paying premium rates for something predictable. Hire when the function needs deep context about your customers — context an outside partner would have to rebuild every engagement. Hire when responsiveness matters more than polish, because internal people answer in minutes and vendors answer within their SLA. Hire when the knowledge should compound inside your business instead of leaving with a contract.
And hire when this is the capability you intend to compete on. Whatever you want to be known for, own it.
Outsourcing buys output. Hiring buys capacity and context. If the work is steady, close to your customers, and something you could evaluate well, you're buying capacity. If it's lumpy, specialized, or definable on paper, you're buying output — and paying a premium rate for it is usually the cheaper mistake.
The Middle Zone, and How to Break the Tie
Most real decisions aren't clean. The volume is almost enough. The work is somewhat specialized. You could write down half the specification.
When you're genuinely torn, outsource first — deliberately, as a way of learning the job. Six to twelve months with an outside partner teaches you what the work involves, what good output looks like, how long it actually takes, and what you'd need to ask a candidate. You come out of it able to write the job description, evaluate the hire, and manage the role. That's a real education, and it's cheaper than a bad hire.
This sequencing is why it's the right default for the middle zone: outsourcing is reversible and hiring is not. Ending a vendor agreement costs a notice period. Ending an employment relationship costs severance, morale, your time, and the confidence to hire again for six months. When two options are close on the merits, take the one you can undo.
Then set the trigger in advance, while you're still calm: when this function reliably takes more than X hours a month, or spend crosses what a fully loaded employee would cost, we revisit hiring. Written down, that's a plan. Left unwritten, it becomes the thing you notice two years late.
The Failure Modes Worth Naming
Three patterns account for most of the regret.
Outsourcing chaos. Handing a broken process to an outside partner doesn't fix it; it adds a contract and a communication lag to the same mess. Providers optimize what you hand them. Hand them disorder and you get expensive disorder.
Hiring to escape a decision. When a function is failing and you can't name why, hiring feels decisive. It rarely is. If you can't articulate what the person will do differently, the role inherits the problem and now it has a salary.
Never revisiting. The most expensive version isn't picking wrong — it's picking once. Owners run a vendor relationship for six years past the point where the volume clearly justified a hire, or keep a role that stopped having a full week of work in it two years ago. Both decisions were right when they were made. Nobody scheduled the review.
Put it on the calendar annually. Ten minutes per function, one question: given today's volume and today's cost, would we make this same call from scratch?
A Thirty-Minute Version of This
You can get to a defensible answer in one sitting. Write the function at the top of a page and work down:
- Hours. How many hours a month does this genuinely require today — not at your best month, at a normal one?
- Both totals. Fully loaded annual employee cost against annual outsourced spend at those hours. Note the gap.
- The four questions. Steady or lumpy? Close to the customer or far? Specifiable or taste-based? Supervisable by you or not?
- Count. Three or four answers pointing the same way means you already have your decision and were looking for permission.
- Split ties toward reversible. Outsource, learn the job, set the trigger that would flip it.
The step that gets skipped is the one where someone who isn't you looks at the page. Owners are systematically bad at estimating hours for work they've been absorbing personally, and systematically optimistic about their ability to manage a discipline they've never worked in. Both errors point the same direction — toward hiring too early — and neither is visible from the inside.
That's the real value of outside perspective on a decision like this. Not someone who knows your industry better than you do, but someone with no stake in the answer who will ask what you'd have to believe for the numbers on the page to be true. It's a fifteen-minute conversation that routinely saves a year.
Frequently Asked Questions
Is outsourcing cheaper than hiring in-house?
Per hour, almost never. Per year, often — and that's the comparison that matters. A provider charges a higher rate but bills only for work performed; an employee costs a fully loaded amount well above the salary line and keeps costing it through slow weeks. Compare annual fully loaded cost against annual outsourced spend at your realistic volume. Most of the time the gap is smaller than expected, which frees you to decide on the shape of the work instead of the price.
What should a small business never outsource?
Whatever customers are actually paying you for, the relationships that would walk out with a provider, and any work requiring judgment you can't write down. If your reputation rests on how a specific thing gets done, that thing belongs to your team. Nearly everything else is fair game — including work that feels important but is performed identically at every business in your industry.
How do I know when to bring an outsourced function in-house?
Three signals have to land together: volume has gone from lumpy to steady, annual spend has climbed toward what a fully loaded employee would cost, and you now understand the work well enough to write the job description and judge the output. All three and hiring converts a variable cost into a cheaper fixed one. Only the third — rising spend — and hiring usually just relocates the problem.
Talking yourself into a hire?
Boule Board gives you a virtual board of directors that knows your business — an outside read on the decision, and the accountability to revisit it when the numbers change.
See Plans →