How to Go From Solo Virtual Assistant to Agency Owner

Agency is a bigger word than the thing it describes. People picture an office, a team page, a name that sounds like a company. What actually separates a solo virtual assistant from an agency owner is smaller and much less romantic: some of the work gets delivered by someone who is not you, and the client still receives what they were promised. Everything else is decoration you can add later.

My own work has been on the delivery end of virtual assistant and BPO accounts, plus the receivables that trail behind them, and I have seen people try this in both orders. The ones who struggled almost always hired first and built the structure afterwards, so their first hire arrived as a training problem stacked on top of an already full week. The move works in the other order, and the part nobody warns you about is that your own job changes more than the work does.

One term before the rest. The person you bring on is a subcontractor: an independent contractor doing work on your clients’ accounts, never on accounts of their own. I will say contractor from here.

Table of contents

An agency is a delivery change, not a branding change

Registering a business name does not make you an agency. Neither does saying “we” in your emails. The change is structural, and there is a plain test for whether it has happened: can a client task be completed to your standard by someone other than you, without the client noticing a drop?

If yes, you have an agency, even if it is you and one contractor working ten hours a week. If no, you have a solo business with a website that overstates it. That gap is where the reputational damage lives, because you have sold capacity you cannot actually deliver, and the first busy month proves it.

Which turns the whole thing into one repeated question. What would have to be true for someone else to do this well?

The three tests that say you are ready to hire

Read the answers off your calendar and your invoice history for the last sixty days, not off how worn out this particular week has left you. Exhaustion is a bad signal because it is always available.

The first test is whether you turned work away. A referral you did not chase, an expansion request from a client you already have, a project you would have said yes to a year ago. Once room in your week is what decides your revenue, capacity has become your ceiling.

The second is whether five hours a week or more are going to tasks you would happily buy at a fraction of your rate. First-pass email, draft listings, entering data, tidying formats, the monthly report you rebuild by hand. Each of those hours quietly bills you the difference between your rate and the going rate for the task, and it does it again the following week.

The third is whether a single quiet week turned up in the last three months. A slow afternoon does not count. A whole week that ended with room to spare does. With no slack at all, a fever, a client emergency and a new account all arrive as crises, and one of them is always on its way.

Two of the three coming back yes means hire. Only one usually means the problem is your process rather than your volume. From inside a busy week those two feel identical, which is why owners so often hire against the wrong one and end up paying more for the same mess. When you cannot tell which you have, audit the process first and hire into one that works.

Define the slice before you look for the person

The most common first-hire mistake is advertising for a virtual assistant instead of defining a job. What you need written down before anyone applies is a slice: the specific tasks, the hours per week, and what finished looks like for each one.

Without it, what you get is someone standing by for instructions, and every instruction still routes through you, which is the exact thing you were trying to escape. You also cannot judge them fairly, because neither of you ever agreed what good would look like.

This is also the point where documentation stops being optional. If the task lives only in your head, handing it over means narrating it out loud every time, and you will quietly decide it is faster to keep doing it yourself. It is faster, for one more month. Writing the SOPs for your virtual assistant business before you hire is what turns a second person from a cost into capacity.

Screen with one paid test task

Interviews reward people who are good at interviews. A small paid test rewards people who are good at the work, and they are not always the same people.

Pay for it. Skipping the fee does not screen for commitment. It screens for a lack of alternatives, so the people worth hiring are the first to drop out. Paying also tells a candidate, before day one, exactly what sort of client you are. One hour, at whatever rate you plan to offer, is enough.

Build the test out of the real queue rather than inventing something clean. If the slice is inbox and listings, send twenty sample emails to triage with one line of reasoning each, plus one listing drafted from a short brief. Give it a 48-hour window.

A pass reads like this. Your instructions were followed as written. The decisions match yours, or the reasoning behind them holds up when you read it. One or two sensible questions came back instead of silence on the parts you left ambiguous. A candidate who says plainly which bits were unclear has the habit that matters most in this work, and it is not one you can train into someone later.

The failures tell you as much as the passes. Delivery late, with no message. Guessing instead of asking. Quietly reshaping the brief into the job they would rather have done. None of that improves after an offer. Whatever the test shows you is the high point, not the starting point.

Paper before the first hour

No signed agreement, no start. You would not open a client engagement on a promise, and the standard does not loosen when you are the one hiring. The document itself does not need to be intimidating to do its job. It needs to cover the delegated scope, the rate and the payment schedule, a paid trial with a short notice period on either side, confidentiality that survives the end of the engagement, and non-solicitation, which is the clause saying your contractor will not approach your clients directly to work for them.

The trial mechanics are worth being specific about. Thirty paid days that either side can end on three days’ written notice keeps a bad fit cheap to leave, and a longer notice once the trial is passed is what passing it earns. Fourteen days is a reasonable figure there. Write all three numbers into the document rather than leaving them as an understanding.

Access is the other half of the paper, and it is the half people improvise. Everything your contractor works inside should be a seat you granted and can take back, under logins you hold. They do not get your client’s credentials, and they never get yours. The test for any access decision is whether you could switch it off in a single action, with no password resets and nothing the client would ever see. If the answer is no, restructure it until it is yes.

The first thirty days decide it

Book fifteen minutes at the end of each week and write down three things: the work that landed, the corrections you had to make, and anything your contractor had to ask about. By day 30 you will have four short entries, and they will make the decision for you. Without them you will judge the whole month on whichever week is freshest in your memory.

Let the checking taper as evidence arrives. In week one you look at everything, which is calibration rather than suspicion, and week one is the only point where checking every item costs you almost nothing. Week two moves to spot checks. By week four you should be reading exceptions, not output. If that taper never starts to feel safe, do not treat it as a process you have not finished building. Treat it as your answer.

Keeping someone past day 30 looks like work landing on time, corrections thinning out week over week, questions turning up ahead of mistakes rather than after them, and the work you handed over no longer sitting in your head. That last item is why you did this in the first place.

When it is clearly not working, close it before day 30 rather than after. Give the notice the agreement calls for, pay through the final day, and keep the message brief and decent. A failing trial that gets extended was extended out of guilt, and it costs both people more than the honest version would have.

What changes about your job

Delegating the work does not make your week emptier at first. It makes it different, and the new work is owning the client.

Three boundaries hold from the first day, however well the trial goes. Money talk stays with you, because a contractor answering a rate question is making a commitment only you can honor. Anything the client will read passes your eye first, and it keeps doing so until you consciously change that. And you remain the person the client deals with: the contractor is behind the work, the account is yours. That last boundary keeps the client’s trust intact, keeps your business out of avoidable trouble, and spares your contractor from decisions their brief does not cover.

Your margin is now the gap between what you bill and what you pay, which makes checking the work part of the product rather than paperwork around it. When something goes out with an error, the client does not see a contractor. They see you. Managing that well without hovering is its own skill, and managing a remote virtual assistant team is the next thing to get right once the first hire has passed the trial.

Do not start this yet if

Three situations make this move worse rather than better, and all of them are common.

The work is not documented. You will spend the trial narrating instead of delegating, decide that hiring does not work, and end up back at a full week having also paid someone.

Your money is not moving reliably. If invoices land whenever clients feel like paying them, a wage bill falling due every two weeks is a cash flow problem you built on purpose. Get collections predictable first.

You are hiring to escape a client you should have ended. Handing a difficult account to someone with less authority than you fixes nothing. It just puts a person between you and a decision you have been avoiding.

If you are unsure whether the foundations are there, a virtual assistant business rather than a job you cannot leave covers the three systems worth having in place before you hire anyone. And if the client lifecycle itself is still improvised, start further back with managing clients as a virtual assistant.

The version that works

The owners who make this move successfully are not the boldest ones. They are the ones who wrote the slice down, tested one candidate with real work, put the arrangement on paper before the first hour, and reviewed properly for a month while the stakes were still small. It is an unexciting sequence, and it is the whole difference between an agency and a solo business carrying an extra expense.

If you would rather start from finished documents than build the process from scratch, the hiring and onboarding SOP, the contractor agreement, and the process audit that tells you whether your problem is capacity or leakage are all part of the Virtual Assistant Business Owner Full System.

Frequently asked questions

What counts as a virtual assistant agency? The working definition is delivery, not size or branding. Once client work can be completed to your standard by someone other than you, and the client still gets what they were promised, you are running an agency. One contractor at ten hours a week qualifies.

When should I hire my first subcontractor? When at least two of three things held true over the last sixty days: you turned work away, five hours or more of your week went to tasks worth a fraction of your rate, and no genuinely quiet week turned up. Only one of the three usually points at a process problem rather than a capacity one.

Should I pay for a test task? Yes. Leaving it unpaid screens for a lack of alternatives rather than for skill, so the candidates worth hiring are the ones who walk away. An hour of paid work at your intended rate is enough to see how someone handles instructions and whether they ask questions.

Do I need a written agreement with a subcontractor? Yes, signed before any work starts. It should cover the delegated scope, rate and payment schedule, a paid trial with a short notice period, confidentiality that outlasts the engagement, and non-solicitation of your clients. Nothing about the arrangement should rest on memory.

Should my contractor talk to my clients? Not in the early weeks, and never about money. Through the trial, everything the client reads should pass your review first, and rate or invoice questions stay with you for good. The relationship is the thing you are actually protecting.