The Real Cost of a Bad Admin Hire, With the Arithmetic

What a wrong administrative hire actually costs a small business, item by item, and why the cheapest fix is ten minutes before the offer rather than three months after it.

Most owners of small businesses can tell you what a bad admin hire felt like. Very few can tell you what it cost, because the cost never arrives as a single invoice. It arrives as an hour here, a lost client there, a Saturday spent redoing a week's filing, and a general sense that something which was supposed to save time is consuming it. This page puts numbers on that feeling, so you can decide for yourself whether a ten-minute skills check before hiring is worth the trouble. Spoiler: the arithmetic is not close.

The four places the cost hides

1. Salary for work that was not done

The obvious cost, and the smallest. If an assistant on $1,200 a month produces, in effect, half an assistant's output because the other half has to be redone, you are paying $600 a month for nothing. Over the three months it typically takes an owner to admit a hire is not working, that is $1,800. Painful, but if that were the whole story nobody would bother screening.

2. Your time checking and redoing

This is the real number, and it is the one owners under-count. An assistant who cannot be trusted with a three-step instruction has to have every output checked. Checking is not free. If you spend forty-five minutes a day reviewing, correcting and re-explaining, that is roughly fifteen hours a month. At any plausible value of an owner's hour, $75, $150, $250, that is $1,100 to $3,750 a month, every month the hire lasts. Over three months: $3,400 to $11,000. And this is time that was supposed to be freed up. The whole reason you hired was to get those hours back.

3. The mistakes that reach the outside world

A wrong date on a client appointment. A supplier paid the wrong amount. A proposal sent with last quarter's figures. Each of these has a direct cost (the refund, the apology discount, the afternoon on the phone) and an indirect one (the client who quietly decides you are disorganised). We have never met an owner who could put a precise number on this, and we will not pretend to either. But we have met several who lost a client worth more than a year of the assistant's salary to a single avoidable error. Call it a tail risk: usually small, occasionally enormous, always present when basic skills are missing.

4. Hiring again

When the hire fails, you are back at the start: writing the advert, reading applications, interviewing, onboarding. If the first search took you twenty hours, the second takes twenty more, plus the weeks of lost productivity in between. Owners who hire offshore through agencies also pay a placement fee again. Budget another $1,500 to $3,000 in time and fees.

Adding it up

For a $1,200-a-month assistant who turns out to lack the basics and lasts three months:

| Cost | Low | High | |---|---|---| | Salary for undone work | $1,800 | $1,800 | | Owner time checking and redoing | $3,400 | $11,000 | | Mistakes reaching clients and suppliers | $500 | $10,000+ | | Hiring again | $1,500 | $3,000 | | Total | $7,200 | $25,800+ |

Against that, a skills screen costs ten minutes of the candidate's time, ten minutes of a supervisor's, and a few dollars. Even if screening only catches one bad hire in five, it pays for itself many times over on the first one it catches.

Three worked examples

Numbers in a table are abstract. Here are three patterns we see constantly, with the cost each one produces.

The careful-sounding assistant who cannot hold a list. Interviewed beautifully. Took notes on every call. Still missed one item in three whenever an instruction had more than two parts, because the notes were never complete and the gaps were filled from memory. Lasted four months. Owner's estimate of time spent re-checking: an hour a day. Cost on our table: around $12,000 before the replacement was found.

The fast assistant who never noticed errors. Quick, pleasant, and sent everything out immediately. Including the proposal with last year's prices, twice. The second client did not come back. Direct cost unknown; the owner put the lost account at roughly $8,000 a year. The assistant's salary over the same period was less than half that.

The assistant who could do everything except follow the exception. Every standard task done well. Every "do this as usual, except today skip the deposit step" done as usual, including the deposit step. In a business where exceptions are the expensive cases, that is the one skill that mattered most, and it was the one nobody had tested.

In all three cases, a ten-minute work sample with a reordered instruction, a hidden list and a sentence to correct would have shown the gap before the offer. In all three cases the owner said the same thing afterwards: "I had no way of knowing." That was true of the process they used. It is not true in general.

Why these hires slip through

If the cost is this high and this common, why does it keep happening? Because the hiring process for admin roles is almost entirely built from signals that do not measure the job:

• The CV measures the ability to write a CV, or to pay someone who can.
• The interview measures conversation, confidence and likeability. All pleasant. None of them is the job.
• References measure whether the candidate can find two people willing to say nice things.
• The agency's "vetting" usually measures typing speed, an English test and a personality questionnaire, none of which asks the candidate to follow a multi-step instruction and act on it from memory.

The thing you actually need, the ability to follow instructions exactly, keep track of details and notice errors, is never observed. So the hire is a guess dressed up as a process.

The remote multiplier

Everything above applies to any admin hire. Remote and offshore hires multiply it. In an office you would notice in the first week that the new assistant does not write things down. Remotely, you notice when the consequence arrives. The feedback loop that catches a struggling hire early is gone, so the three months of cost above can become six. Time-zone gaps slow every correction. And the competitive labour markets that make offshore hiring affordable also push candidates to present themselves as more experienced than they are, because the next applicant will. None of that is a character judgement about any country or any person. It is what the incentives produce, and the employer's only defence is to check rather than to trust.

What checking looks like

A ten-minute supervised work sample. The candidate, on a live video call with someone from your team, carries out a short sequence of everyday admin tasks: follow a four-step instruction with one step moved, remember three day-and-task pairs after they disappear, correct a sentence with the wrong weekday, add two figures, apply a simple if-then rule. The values are freshly randomised for every attempt so nothing can be shared. The candidate's camera and screen are recorded. The result arrives as a score, a plain band and an item-by-item breakdown, and a person on your team decides what it means.

That is what the Admin Skills Screen is. It will not tell you whether a candidate is a good person, whether they will be happy in the role, or whether they will stay. It will tell you, before you pay the first month's salary, whether they can do the small concrete things that the first month requires. Given the arithmetic above, that is the single highest-return ten minutes in the whole hiring process.

What to do with the time you save

The point of an admin hire is to buy back the owner's hours, so it is worth closing the loop. If screening removes even one wrong hire from your next five, the table above says you keep somewhere between seven and twenty-five thousand dollars' worth of time and money. That is not an abstract saving. It is the difference between an assistant who lets you stop checking and one who makes you check more.

The employers who get the most from screening use the result to shape the first month rather than just the offer. A candidate who scored well across the board gets autonomy early. A candidate who scored well except on the multi-step items gets written checklists for anything with more than two steps, and a review point at the end of week two. The screen tells you where to look. Looking there, in the first fortnight, is what turns a decent hire into a good one.

Why owners under-count it

Three habits of mind keep the cost invisible, and naming them helps.

Sunk-cost patience. Having spent twenty hours hiring, the owner wants the hire to work and extends the benefit of the doubt for a month longer than the evidence warrants. Every extra month is another row in the table above.

The "I would have had to do it anyway" fallacy. Time spent checking an assistant's work feels like time you would have spent doing the task yourself. It is not. Checking is in addition to the salary, and a reliable assistant removes it entirely. The honest comparison is not "assistant versus no assistant"; it is "this assistant versus a reliable one".

Attributing mistakes to the situation. The first wrong date was the client's confusing email. The second was a busy week. The third was the time zone. By the fourth, the owner is explaining a pattern as a series of accidents. A skills screen, run before the hire, replaces that chain of excuses with a single observed fact about whether the person can hold a date.

What the cost looks like at scale

A business with one assistant feels the cost as the owner's lost evenings. A business with five feels it as a manager's entire job. We have spoken to small firms where one person's role had quietly become "check everything the assistants send out", a full-time salary spent compensating for a hiring process that never looked at the one skill that would have made the role unnecessary. At that scale, the arithmetic above multiplies by the number of hires, and screening stops being a nice-to-have and becomes the cheapest line in the budget.

A note on using the result responsibly

A screen is a reason to look closer, not a reason to stop looking. Use it as one input alongside the interview, references and, for important roles, a short paid trial task. Offer extended time to anyone who asks for it. Never let a score be the sole reason for a decision, and never let software make the decision for you. Those are not just good manners; they are what keeps a screening programme on the right side of employment law, and they are built into how this product is designed to be used.

Frequently asked questions

What if a strong candidate has a bad day? It happens, and it is why the result is one input. If someone you rate highly scores poorly, look at which items they missed and ask about them. Sometimes the answer is a connection problem or a misunderstanding of the format. Sometimes it is exactly the gap the screen exists to find.

Is ten minutes really enough? For the specific skills being checked, yes. The tasks are short because the skills are basic. Longer tests mostly add fatigue.

Should I screen existing staff? Many employers run it on the current team first, both to see what "good" looks like in their own context and to make sure the screen feels fair before candidates take it. It is also a useful, low-stakes way to set a benchmark.