Virtual assistant pricing and contracts: hourly or monthly

Do I charge by the hour or by the month?

The short answer: hourly first, monthly after. And there is a reason for that order.

At the start of an engagement neither of you knows how long the work takes. The client thinks answering their page's messages is half an hour a day and it turns out to be two hours. If you fixed a monthly amount at that moment, one of the two of you has lost, and the relationship goes wrong in month one. Hourly billing splits that risk: the client pays for what actually happened and you are paid for the time you actually spent.

But hourly has one large flaw that is rarely said out loud: it penalises you for getting better. Work that took three hours in month one takes forty minutes in month three, because you learned it and built yourself a pattern. Under the hourly model, the very skill that raised your value has lowered your income. It is not the only flaw of hourly, but it is the one that comes for every assistant who genuinely gets good.

A fixed monthly amount, also called a retainer, solves that. The client pays for a defined result in the month rather than for a number of hours, and any speed you gain is yours to keep. In exchange it carries two conditions, and without them it becomes a trap: the work must be repeating and definable, and the ceiling must be written down. "Whatever comes up" is not a monthly agreement, it is full time employment at half the pay.

There is a third model and for some work it is the best one: a price per unit of work. Per post, per invoice, per article typed. It is right when the unit really is identical. If a three line invoice and a hundred row invoice carry the same price, you will simply start waiting for the big ones to arrive.

One thing said plainly: none of these three models fixes a wrong rate. If you worked the floor number out incorrectly, hourly or monthly makes no difference. The next section is about that number.

How do I work out my own rate?

From the bottom up, and out of your own numbers. You will see no figure here in Toman or dollars, and that is deliberate: no citable rate has been published for this market, and a number we invent would be both wrong and misleading six months from now. The method, though, does not go stale.

One. Write down your monthly cost. Not just living money: internet, electricity, the monthly share of hardware that eventually has to be replaced, tool subscriptions, and money set aside for tax and for sick days. Someone who does not count those last three is subsidising the client out of their own capital.

Two. Count sellable hours, not working hours. This is where nearly everyone goes wrong. If you work eight hours a day you do not have eight sellable hours. Finding clients, answering messages that never become work, writing invoices, learning a new tool, and rework you are not paid for all come out of those eight. Do not guess the real number; log it for one week and then count.

Three. Divide. Monthly cost divided by sellable hours in the month is your floor. The floor, not the price. Working below it means every hour you sell moves you backwards, and that has nothing to do with the market or with competition; it is plain arithmetic.

Four. Go from the floor to a price. The price is the floor plus what the work is worth to the client. Work that only frees up a client's time and work wired directly into sales carry two different prices, even when they take you the same amount of time. An assistant answering a shop page's sales messages is standing in the path of the money, and can say so.

And if your floor comes out above what this market pays? The answer is not to go below the floor. It means this combination of work does not add up and something has to change: the work gets more specialised, or the clients get more numerous and the tasks smaller, or the costs come down. Selling below your floor only postpones the date of the failure.

The four things that actually move a rate

The hub of this path promised that this lesson would open these four up. Here they are, in the order in which they matter most in practice.

One. Hourly or per task. The same argument as the first section, seen from the rate: under hourly billing your income ceiling is your hours ceiling and it does not rise past a point. Under per task and monthly models the ceiling is set by how efficient you are rather than by the hours in a day. This is the only place where your effective rate can multiply without you working more.

Two. How many clients at once. The effect runs both ways and more is not always better. Several clients means safety, because one leaving does not knock you over. But switching costs something: every move between two clients takes minutes to get back inside their world, and someone carrying six small clients at once spends a noticeable part of the day on that switching alone. The right number differs per person and is discovered, not copied.

Three. Responsibility for money and access. This is the biggest jump in rate and the least discussed. An assistant who only answers messages and an assistant with access to the payment panel, the customer list or the business's main email address are doing two different jobs. The second carries risk, and risk is priced. If that level of trust has been given to you, the rate should show it; and at that same moment your confidentiality undertaking should be written down, which is the next section. The security basics lesson takes up limited, revocable access separately.

Four. Whether the money comes from inside Iran or outside. Rates in those two markets are not comparable at all, but working with a foreign client brings two problems that have nothing to do with assistant skill: language, and getting the money in. The second is not simple for Iran and we do not solve it in this lesson; the payments path follows that subject on its own.

There is a fifth factor that was not on the list and should be said: a record you can show. Two people with identical skill, one with three real testimonials and one with none, get two different rates. That is not unfair; the client is paying to reduce risk and you are the one reducing it.

What does a minimal agreement put in writing?

Five things. And before anything else, a warning is needed: what follows is practical guidance and not legal advice. The type of working relationship, tax, insurance and the legal force of any undertaking are the business of someone whose job that is. What we are saying is which things, left unwritten, become the argument later.

Scope. The list of tasks inside this agreement and, more importantly, one sentence about what is not in it. "Answering Instagram messages" is not a scope; "answering Instagram messages during the agreed hours, except technical questions, which are passed to you" is a scope. Without that sentence the work grows over three months and nobody can say where it grew.

Hours and the response window. How many hours a month, on which days, and above all: how long after a message a reply arrives. If that is not written, the client's default expectation is "right now", and you never see that expectation until the day it goes unmet.

Confidentiality. This is the undertaking the two previous lessons of this path handed forward to this one. It does not need to be legal text; it needs to be definite. Three plain sentences: what the client's information is (their customers, their prices, their messages, anything not public), what you will not do with it (not take it elsewhere, not put it in a tool the client has not approved, not show it to anyone), and what happens at the end (your copies are deleted and your access is closed). Most people drop that last clause, and it is the one that matters in practice.

Payment timing. Not only the amount: when. "End of each month", or "seven days after the invoice is sent", or "half at the start, half on delivery". For a new client, asking for money up front is not a discourtesy; it is what turns the start into a decision and identifies the party who was never serious.

Ending it. Both sides must be able to end it and how must be written down. A notice period, two weeks say, for both sides. That clause is in your favour rather than the client's: without it, a monthly agreement cut off suddenly means that month's income is gone and you had no chance to fill the gap.

And the form of this agreement? One written message in the same place you work together, confirmed by the client in a sentence, beats a ten page contract nobody read. What has value is that it was written down, not that it was formal.

Are you selling assistance or availability?

These two are not the same thing, and the fourth lesson of this path left it open. We settle it here.

When you say "I do your admin work", you are selling work. When you say "message me any time and I reply within the hour", you are selling something else: a commitment to be present. Those two products differ in every respect.

Work gets cheaper to deliver as you get better. Availability never gets cheaper, because what it costs is calendar time and you cannot get faster at that. However skilled you become, the window in which you must keep an eye on the phone is the same width. Which is why pricing availability by the number of tasks is wrong; it has to be priced by the width of the window.

Our position: either do not sell availability, or sell it explicitly and separately. If a client genuinely needs fast replies, that is its own line item with its own price and its own hours, not something attached free to a monthly agreement. And if you are not selling it, write your response windows into that same agreement; the two or three fixed windows a day from the skills lesson.

There is a trap here specific to people who have just got fast: when better tools and a better workflow let you finish the work in half the time, you are tempted to sell that speed as an advantage and say "I am always available". That moment is precisely where an independent business turns into an employee without an employment contract. Getting faster lets you take on more clients or finish earlier; it does not licence you to give your calendar time away.

When and how do I raise my rate?

There are three situations where raising the rate is natural and easy to explain: when the scope has grown, when the agreed period ends, and when your own costs have risen. Outside those three it is possible, but harder to explain.

Three rules make it bearable for both sides. First: give notice in advance, not attached to an invoice. Give the client a month to decide, because deciding and being subjected to something are two different experiences and later the client remembers only the experience. Second: never go backwards; the new rate applies from a stated date and work already done is billed at the old one. Third: give a reason, but a sound one. "The work went from three items to eight" is a reason. "Inflation" is a weak one because you have no control over it and neither does the client; if that is genuinely your only reason, say honestly that your own costs have risen, and that is enough.

Some clients will leave. That is not a sign you got it wrong, it is a sign the mechanism is working: the new rate filters out clients for whom the work is not worth that much. What matters is having spare capacity for the next client at the same time as you announce it. Raising your rate while your book is full is the least risky version of it.

There is one common mistake worth naming: holding an old client's rate down as a loyalty discount. In practice the result is that you do your hardest work at your lowest rate, because it is the old client who sends the most work. If you want to reward loyalty, make the increase smaller or slower, but do not stop it.