Part-Time vs Full-Time Virtual Assistant: Why Hours Are the Wrong Question

Choosing between a part-time vs full-time virtual assistant usually starts with counting hours. Then this happens: you bought a package of hours, there are plenty left this month, and the urgent request from this morning is still sitting there, untouched.

Your assistant isn't slacking. They're just not online right now, because you bought time, and what your customer needed was someone available at a particular moment.

That's the trap at the heart of the part-time vs full-time virtual assistant decision. People think it's a question about how much work there is. It's really two questions tangled together:

1. How much work is there?
2. When does it need doing?

Untangle those, and the right arrangement usually becomes obvious. Leave them tangled, and you'll either pay for a full-time assistant who's twiddling their thumbs, or buy a part-time package that keeps missing the moments that matter.

This guide shows you how to work out what you actually need, including the maths that quietly catches most people out.

Here, a virtual assistant (VA) means an independent freelancer or agency you outsource work to. "Part-time" and "full-time" describe how much service you're buying, not employment status.

A small workload can need a big presence

Here's the counter-intuitive bit.

Say customer emails take about 30 minutes a day in total. Tiny workload. A part-time VA could handle that in no time.

But those emails arrive all day long, and your customers expect a reply within a couple of hours.

So now you don't need 30 minutes of help. You need someone keeping an eye on things across your whole working day.

That's coverage, and it's a completely different purchase from workload.

The reverse also happens. A big weekly report might take eight hours. Sounds like a lot. But if it just needs to be ready by Monday morning, it can be done any time over the weekend, in any time zone.

Big workload. Tiny coverage need.

Most bad VA arrangements come from buying one when you needed the other.

Draw the arrival pattern before you choose a package

Before you look at any pricing page, look at your own work.

Open wherever requests arrive: your inbox, order system, booking form or project board. Pick a normal week. For each request, note:

• When it arrived.
• When it genuinely needed a response.
• How long the actual work took.

Keep waiting time out of that last number. If you sent a supplier a question and waited a day for an answer, that's a day of waiting, not a day of work.

Now step back and look at the shape. You'll usually see one of these patterns:

• Clustered. Lots of work right after a delivery run or a launch, quiet otherwise.
• Even drip. A steady trickle through opening hours.
• Deadline-driven. Work that piles up before one weekly or monthly cut-off.

A weekly total of hours hides all of that. Two businesses with "ten hours of admin a week" can need completely different arrangements.

Be honest about what's really urgent

When you ask yourself how quickly each request needs handling, you'll be tempted to say "straight away" for everything.

Resist that. Ask instead:

• Is this a promise we've made to customers?
• Is it a customer expectation we'd hurt by ignoring?
• Or is it just my habit of replying instantly?

A document needed for tomorrow's meeting can easily wait for an afternoon batch. A same-day cancellation that frees up a technician for someone else probably can't. It loses its value if nobody sees it until tomorrow.

If you mark everything urgent, your VA can't tell a real emergency from a routine job. And you'll end up buying expensive coverage for work that could have waited.

Pick out the handful of request types where timing actually changes the outcome. Design around those. Let everything else go in batches.

Count active work, not open tickets

Here's where estimates usually go wrong in one of two directions.

Mistake one: counting the waiting as working. A case might stay open all day while you wait for a customer to send a missing document. But the actual work might have been two quick emails. Don't plan as if open cases take continuous effort.

Mistake two: counting only the typing. Sending the final reply might take a minute. But reading the request, checking the order history, confirming the policy and updating the record are all part of doing it properly. Leave those out and your estimate is far too low.

Time a realistic sample of tasks. Label each figure clearly as "measured" or "guessed."

And remember: your own speed is only a starting point. Someone learning your systems will be slower at first. Leave room for that.

Three different promises hide inside "available"

When someone says they'll be "available," they could mean three very different things:

• Coverage: someone is around to notice and handle requests during an agreed window.
• Response time: how quickly something gets acknowledged after it arrives.
• Completion time: how quickly the actual result is ready.

An acknowledgement ("got it, looking into this") isn't the same as a resolution. If you need a booking confirmed, not just seen, say so.

And no provider can promise a completion time if they're waiting on you for a decision. Build that in.

Also ask the awkward question: what happens if your assistant is busy with one of your tasks when another urgent one arrives? Is there a queue? A backup person? Or does it just wait?

The answer tells you whether you're buying one person's time or a service with some resilience built in.

Match the service pattern to the timing, not the job title

Once you know your workload and your timing, you can pick an arrangement that fits. There are more options than "part-time" or "full-time":

Flexible hours with deadlines. Great for reports, research and anything that just needs to be done by a certain time. The VA organises their own schedule around your deadlines. It's not an invitation to expect an instant reply whenever you remember something.

Scheduled processing windows. For example, your VA clears the queue at 10am and 3pm your time, with a separate route for genuine emergencies. Brilliant for steady drips of routine requests.

A retainer. A monthly payment that reserves a mix of capacity and availability. But "retainer" means different things to different providers. Read exactly what it buys.

A dedicated full-time service. Suits sustained demand across most of the week. Even then, it's one person. They can't be in two places at once, and they need breaks and cover.

A one-off project. Perfect for clearing a backlog. Keep it separate from the ongoing flow of new work, because those are two different jobs.

A small, well-placed amount of time often beats a big block of unfocused hours.

Let's do the maths on a real-looking week

Here's a worked example. The numbers are invented to show the calculation, not typical figures.

A business coordinates equipment repairs. In a normal five-day week, it gets 30 complete requests. Each takes about 12 minutes of active work: checking availability, confirming customer details, preparing the booking and updating the record.

30 × 12 minutes = 360 minutes, or 6 hours.

Add the rest:

• 2 hours of record maintenance
• 1 hour of routine follow-up

Total: 9 hours of work a week.

That's nowhere near full-time. Buying a full-time assistant because "we're open five days a week" would waste most of what you pay for.

But there's a catch. Requests don't arrive in one tidy batch. Some customers want a quick confirmation before they arrange site access.

So the owner chooses:

• a part-time service with morning and afternoon processing windows, plus
• an urgent route that stays with the owner for anything that can't wait for the next window.

That only works if it matches the promises the business already makes to customers. If most requests genuinely need an instant live response, the owner needs to buy that coverage, change the promise, or use a provider with proper backup.

The 9 hours is still useful. It just isn't the whole picture.

Leave room for the busy week, or the busy week will break you

Now let's stress-test it.

Suppose a busy period pushes requests up from 30 to 45 a week. Same 12 minutes each:

45 × 12 = 540 minutes, or 9 hours of coordination.

Add the same 3 hours of maintenance and follow-up: 12 hours.

That isn't a forecast. It's a "what if" that makes you ask useful questions:

• Can the provider add hours with notice?
• Do processing windows get longer?
• Which work can wait?
• Who approves the extra cost?

Don't plan your normal week to use every available minute. If every hour is booked, the first missing document or difficult customer pushes everything else back.

You don't need a magic buffer percentage. Look at how much your weeks actually vary, and leave room for that.

And if the busy pattern keeps happening, update the agreement. Treating a predictable rush as an emergency every single time is exhausting for both of you.

Why 80 hours a month isn't the same as 20 hours a week

This one catches almost everyone, and it's worth knowing before you compare quotes.

Months aren't four weeks long.

If you buy 20 hours a week, over a full year that's:

20 × 52 weeks = 1,040 hours a year

1,040 ÷ 12 months = about 86.7 hours a month on average

So a package advertised as 80 hours a month is actually a bit less than 20 hours every week.

That might be fine. But compare quotes on the same basis. If one provider quotes weekly and another quotes monthly, convert them before deciding which is cheaper.

Then ask the practical questions:

• When does the monthly allowance reset?
• If you start mid-month, is the first month prorated?
• Do holidays and closures reduce what you can actually use?

Our virtual assistant cost guide shows you how to compare quotes properly.

Unused hours aren't savings, and rollover isn't always what it sounds like

A bigger package usually comes with a lower hourly rate. That looks like a bargain.

But an hour you paid for and didn't use isn't cheap. It's wasted.

Some providers advertise rollover, where unused hours carry over to next month. That's helpful, but check the detail:

• Is there a cap on how much rolls over?
• Do rolled hours expire?
• Can the provider actually schedule all those extra hours in the month you need them?

Rollover isn't worth much if your assistant can't fit the accumulated work in when you need it.

Also agree how extra work gets approved. You don't want a surprise invoice, and your VA doesn't want to do extra work without knowing they'll be paid.

And ask what happens when your assistant is ill or on holiday. Can cover handle routine tasks? What about sensitive ones? Who keeps your unfinished work visible so nobody's starting from scratch?

You need some overlap, but probably less than you think

Overlap is when you and your VA are both online at the same time. It's useful for training, quick questions and decisions.

But it's easy to overestimate how much you need.

If your VA can prepare a batch of decisions before your morning review, a short, reliable overlap might be plenty. If every single task needs your instant answer, that's a sign your instructions or permissions need work, not that you need more overlap.

Always write times with a named time zone. "Nine in the morning" means nothing to someone on the other side of the world. And double-check what happens when clocks change, because your overlap can shift by an hour without anyone noticing.

One more thing. A green "online" dot in a chat app doesn't prove work is happening. And a quiet chat doesn't prove it isn't. Judge the work, not the status light.

Test with a real deadline before you buy more

Start with a scope that matches your evidence, not your hopes.

Then track a few weeks properly:

• When did each request arrive?
• When did work start?
• What blocked it?
• When was the result ready?

Keep your own delays visible. If a task waited three hours for your approval, that's not the VA's fault.

When a target gets missed, look at why:

• Did it arrive after the cut-off?
• Was information missing?
• Did it need authority the VA didn't have?
• Or was there genuinely too much work?

Only that last one gets fixed by buying more hours. The others need a clearer form, a different window, a backup decision-maker, or better instructions.

Moving to a bigger arrangement makes sense when the evidence supports it. Not because full-time sounds more grown-up.

Write a service window both of you can actually use

Put the practical agreement into a short operating note alongside your contract:

VA SERVICE WINDOW NOTE

Work covered and what "finished" means:

Request volume we measured, and the period:

Estimated active work (and how unsure we are):

Service windows, with named time zone:

Response target vs completion target:

Cut-off time and information needed:

Requests that can wait for the next window:

Urgent exceptions and who decides them:

Hours included and billing period:

Usage warning and how extra work is approved:

Unused hours rule and expiry:

Absence, outage and cover arrangements:

What would make us increase the hours:

Now test it. Imagine a request arriving one minute after the last window closes. What happens? If nobody can say, there's a gap.

Try it with a busy day, and a day when you're unreachable. Those reveal far more than a comfortable weekly total.

The right answer might be a few well-placed hours, a bigger recurring commitment, or separate routes for routine and urgent work. What matters is that the work and its timing fit together, so nobody discovers a hidden availability promise while a customer is waiting.

For the full hiring process, start with how to hire a virtual assistant.

Frequently asked questions

How many hours a week do I need a virtual assistant?

Measure the active work in a typical week, add a buffer for exceptions, then check whether the timing of requests means you need coverage on top. Many small businesses find their real workload is far smaller than a full-time role.

Is a full-time virtual assistant better value than part-time?

Only if you can keep them usefully busy. Bigger packages often have lower hourly rates, but unused hours aren't a saving. Value comes from work completed at the right time, not hours reserved.

Do unused virtual assistant hours roll over?

It depends entirely on the provider and plan. Some advertise rollover with caps or expiry dates. Others don't. Get the exact rule for your plan in writing before buying a larger package.

What is the difference between workload and coverage?

Workload is the time the tasks actually take. Coverage is the period when someone must be available to respond. A small workload spread across the day can need a lot of coverage.

Is a monthly package of 80 hours the same as 20 hours a week?

Not quite. Twenty hours a week across a year averages about 86.7 hours a month, because months aren't exactly four weeks. Compare packages on the same basis before choosing.