Salary is the sticker price, not the cost
When an owner asks whether they can afford somebody, the sum is almost always done against the wage. Forty thousand a year, so I need forty thousand more of margin. That sum is wrong by a wide margin, and the gap is where a lot of small businesses get into trouble in their second year of having staff.
The wage is the one number that is visible, agreed, and easy to plan around. Everything else arrives in pieces: monthly, quarterly, annually, or as a surprise. Employer payroll taxes, unemployment insurance, workers compensation, equipment, software seats, phone, space, training, and your own time managing all of it.
A widely published rule of thumb is that employer taxes, insurance, and benefits add roughly twenty-five to forty percent on top of base wages, before you have bought anybody a laptop or spent an hour training them. The real figure depends on your state, your industry, and what you choose to offer, so treat that as a planning range and get your actual number from a payroll provider or your accountant before you make anyone an offer.
None of this is an argument against hiring. It is an argument against hiring against the wrong number, which is how a hire that felt affordable in a spreadsheet turns into six months of quiet stress.
The load that sits on top of wages
Start with the statutory pieces, because they are not optional and they are not negotiable.
You pay the employer half of Social Security and Medicare, currently 7.65 percent of wages, on top of the employee's own half that comes out of their cheque. Then federal unemployment tax, then state unemployment tax, which varies by state and by your own claims history, so a business that has laid people off pays a higher rate than one that has not.
Then workers compensation insurance, which is the number that surprises people most, because it is priced by class code. The rate for office work and the rate for roofing are not in the same universe. If you are in a physical trade, get a real quote for your specific class code before you make an offer, not after you have already promised somebody a start date.
Then whatever you offer voluntarily: a health contribution, paid time off, a retirement match, a phone stipend. Paid time off deserves more thought than it usually gets, because it is not only money. It is coverage. Two weeks off means two weeks where the work still has to happen and you are the person doing it.
The costs that never appear on a payroll report
These are the ones that make a hire cost more than the spreadsheet promised, because they do not arrive through payroll and so nobody adds them together.
Equipment and setup: laptop, phone, tools, vehicle or a vehicle allowance, uniforms, safety gear, a desk and somewhere to put it. Then a per-seat cost for every piece of software the business runs on, and there are usually more of those than you remember. If you already feel like you pay for too many subscriptions, adding a person multiplies the bill, which is one reason I keep arguing to cut the software stack down before adding anybody to it.
Recruitment: job listings, your hours reading applications, your hours interviewing, background checks, and everything you were not selling or delivering while you did all of that. That cost is real even when the hire works out, and you pay it a second time if it does not.
Administrative overhead: a payroll service, employment practices coverage, an accountant with more to do, record keeping, and compliance you never had to think about when it was only you. Small individually. Not small together.
Training is a cost you pay in your own hours
This is the largest item on the list and the one most owners leave out entirely, because it never arrives as an invoice.
A new person is not productive on day one, and in a small business the person doing the training is the owner, who is also the most expensive and most constrained resource in the company. So the true cost of the ramp is the wage you are paying for output you are not yet getting, plus the output you personally stopped producing in order to teach.
Be honest about the ramp for your particular work. Some roles are useful inside a fortnight. Skilled trade work, anything requiring judgement, anything customer-facing where one mistake costs you an account, takes months. Estimate it, write it down, then add to it, because this is the number everybody underestimates.
The fix is not to train less. It is to have something to train from. If the process only exists in your head, every hire costs you the same full ramp, forever, and so does every replacement. Write the process down before you hire and you pay that cost once instead of every time.
Management is the cost that never ends
Training finishes. Management does not.
Once you have somebody, part of every week belongs to them: questions, checking work, scheduling, feedback, payroll, the awkward conversation you have been putting off, and the morning they call in sick and you cover. In a one-person business that time comes straight out of the work that makes money, because there is nowhere else for it to come from.
This is what catches people who hired specifically to get their time back. A hire does give you time back eventually, but it takes some first, and the exchange rate is worse than expected for the first six months. If you are already at capacity, the hire will make you busier before it makes you freer, and you need enough runway to get through that stretch.
It is also worth asking plainly whether you want to manage people at all. Some very good operators do not, and there is no prize for pretending otherwise. Wanting help is not the same as wanting to be somebody's boss, and building a business around the wrong one of those is a slow and expensive mistake.
What a bad hire actually costs
You will find plenty of confident figures for this. Rather than trusting any of them, do the sum for your own business, because it is not difficult and yours is the only one that matters.
Add up the loaded wages you paid for the months they were there, the ramp you funded and will not get back, your own hours spent hiring and training and correcting and eventually ending it, the work that had to be redone, the customers who had a bad experience and did not come back, and the cost of running the entire hiring process again from scratch. In a small business that total routinely runs to several times the wage, and the biggest line in it is your own time.
Then there is the piece with no number on it, which is what it does to you. A bad hire in a small business is not a line item, it is something you carry around all day. Owners regularly tolerate a bad fit for months longer than they should, because ending it means admitting the whole cost at once and starting over from nothing.
Two things reduce this more than anything else. Hire slower than feels comfortable, and give the work a real trial before it becomes permanent, on a small paid piece of actual work rather than another conversation about work. And define the role precisely enough that both of you can tell whether it is going well, because most bad hires are not bad people. They are unclear roles.
Employee or contractor, and why the answer is not really yours
This is the second most searched version of the question, and it usually gets answered as though it were a preference. It is not. Whether somebody is an employee or a contractor is determined by how the work actually happens, not by what the agreement calls them.
The rough shape: if you control when, where, and how the work is done, you supply the tools, you set the hours, and the person works only for you on an ongoing basis, that relationship looks like employment regardless of what you both signed. If they set their own schedule, use their own equipment, work for other clients, and are engaged for a defined outcome, it looks like contracting.
Getting it wrong is expensive in a specific way: back taxes, penalties, and unpaid benefits, assessed after the fact when you have already spent the money. The rules differ by state and they have been tightening, so this is a question for an accountant or an employment attorney in your state before you hire. It is not something to reason your way through from a blog post, including this one.
Where contracting genuinely fits is defined, bounded work: a project, a specialism you need occasionally, overflow in your busy season. Where it does not fit is 'I want an employee without the employer costs'. That is the version that gets reclassified.
Can I afford to hire? Run this before you post the job
Here is the arithmetic, in the order that keeps you out of trouble.
One: take the wage you would offer and add your loaded percentage for employer taxes and insurance, using a real quote for workers compensation in your class code rather than an average off the internet. Two: add the first-year one-offs, which are equipment, setup, recruitment, and a monthly figure for software seats. Three: add the ramp, valued at the wage you pay during it plus the hours of your own you will spend teaching. That total is the true first-year cost, and it lands well north of the wage you started with.
Now the other side of it. How much additional gross profit, not revenue, does this person have to generate to cover that? Divide the true cost by your gross margin and you get the extra revenue the hire has to produce. If your margin is thirty percent, every dollar of cost needs more than three dollars of new revenue standing behind it. Look at that number and ask honestly whether the demand for it already exists, in work you are currently turning away, or whether you are hoping it shows up once the person starts. Hiring into hope is the version that goes wrong. And if you are not confident your prices support this at all, that is a pricing problem before it is a hiring problem, and I worked the method through in how much should I charge for my services.
Last, and this is the one that actually kills businesses: cash, not profit. Wages are due on a schedule that does not care whether your customers have paid you. Before you hire, know how many months of full payroll you could cover if revenue stopped tomorrow, and know how long your invoices genuinely take to get paid rather than how long your terms say they should. Businesses rarely fail because the hire was the wrong person. They fail because payroll landed on a Friday and the money landed the following Tuesday.
Some work should be a person, and some work should be a system
Here is where I will be straight with you, because this is the point where somebody selling software normally starts exaggerating.
Software does not replace people. It replaces tasks, and the difference between those two things is judgement. Anything that needs reading a situation, handling somebody who is upset, making a call on incomplete information, doing skilled physical work, or holding a relationship, that is a person. It has always been a person and it will keep being a person, and pretending otherwise is how you end up with a fully automated business that customers quietly leave.
But a large share of what fills a small business day is not judgement. It is the same sequence run again: taking the same details off every enquiry, sending the same confirmation, chasing the same unpaid invoice, updating the same three places when one thing changes, writing the same status update, moving information from where it arrived to where it needs to be. None of that needs a person. It needs a decision made once and then executed every time without anybody remembering to.
So the honest test is not about cost. Ask what fraction of the role you are about to hire for is judgement and what fraction is repetition. If it is mostly judgement, hire, and do the sum above properly. If it is mostly repetition, you are about to pay a wage plus a load plus training plus management for something that should have been built once. And if it is genuinely both, which it usually is, systematise the repetitive half first and then hire for a far better job than the one you were going to advertise.
How I run a company on my own, and what that actually means
I am one person. There is no team behind this, no account manager, no bench. That is not a hardship story, it is a design decision, and it only works because the repetitive parts of the business are built rather than staffed.
The systems I built for myself are the ones I now build for other people. A system I built for a real California field-services company has coordinated 705 jobs end to end: intake, scheduling, client updates, invoicing, and the paperwork that follows, with nobody sitting in the middle of it doing data entry. For another client, organic traffic went up 60 percent in 28 days with no ad spend, because the thing that produces it runs whether or not anyone remembers to do it that week.
What I did not automate is the work that needs me: the judgement calls, the difficult conversations, the decisions where being fast and wrong costs more than being slow and right. That split is the whole thing. I am not doing less work than an owner with staff. I am doing only the work that is genuinely mine to do.
If you are looking at a job posting and a payroll estimate and something about it is not sitting right, it is worth spending an hour on that split before you spend a year on the hire. Related reading: what an AI employee actually does and running the whole business on autopilot rather than just the phone. Or send me the list of what fills your week and I will tell you which half is a person and which half is a system.