Every bookkeeper who hits a wall on payment posting gets handed the same two options. Hire a part-time VA, or find some tool. Most people pick the VA because it feels safer. A person you can train beats software you have to trust. I get it. But nobody actually runs the numbers, and the numbers are the whole point.
This one is bookkeeper-to-bookkeeper. I'm not selling you a dream where automation replaces a human and you sip coffee while robots do your books. I'm going to show you the real cost of each path for one specific job: getting payments out of remittance documents and into QBO. That's it. Keep the comparison narrow and the math stays honest.
Key Takeaways
- A part-time VA for payment posting realistically costs $700 to $900 a month before you count your own management time. Automation for the same slice of work runs under $100 a month.
- The two are not the same scope. A VA does many things a tool cannot. The honest comparison is only for the repetitive document-to-QBO step, where 62% of AP processing cost is labor (APQC via ResolvePay, 2025).
- The hidden cost of a VA is not the hourly rate. It is the retraining, the review, and the error rework that follow every new client and every format change.
- The right answer for most practices is both. Let the tool handle the data entry. Let the human handle judgment.
What does a part-time VA actually cost?
A part-time bookkeeping VA for payment posting realistically lands between $700 and $900 a month, and that is before your own time enters the equation. US-based bookkeeping VAs commonly bill $15 to $25 an hour. At nine hours a week and $18 an hour, you are at roughly $700 a month in direct cost. The virtual assistant market is projected to pass $19 billion by 2025 (Grand View Research, 2023), so the supply is there. The price is not the problem.
The problem is everything stacked on top of the hourly rate. You pay for onboarding every new client, because payment posting is not one skill. It is a different skill for every QBO file and every remittance format. You pay for your own review time, since a VA who does not know a client's history will miss edge cases you would catch on sight. And you pay for rework, because manual invoice processing carries an error rate of about 1.6% per invoice (ResolvePay, 2025, citing Turing IT Labs). A misapplied $200 payment does not announce itself. It waits for reconciliation, then costs you 30 minutes to unwind.
From my practice: When I priced out a VA years ago, I stopped at the hourly figure and thought $700 sounded fine. What I did not budget for was the two hours a week I would spend answering questions and checking work. That is not a side cost. That is the job slowly becoming managing someone else doing the job.
What does the automation actually cost?
Automation for payment posting costs a flat monthly fee that does not move when your client count does. Ground Control runs under $100 a month, and that number stays put whether you are posting payments for three clients or fifteen. There is no hourly meter, no onboarding fee per client, and no payroll admin. You connect a QBO file once and the routing handles itself from there.
The honest part is what the fee does not cover: your judgment. The tool extracts invoice numbers and amounts from a PDF, CSV, or image, matches them against open invoices in QBO, and hands you an approval screen. You still look at every batch before it posts. But "looking at a batch" is minutes, not the two hours a week you spend managing a person. The labor cost collapses because the labor is mostly gone. That matters when 62% of total AP processing cost is labor (APQC via ResolvePay, 2025). Cut the labor, cut the cost.
Automated AP operations process invoices at roughly six times the throughput of manual ones, reducing per-invoice cost by 70 to 80% (IOFM via ResolvePay, 2025). I am not going to pretend my numbers match an enterprise AP department's. But the direction is the same one I lived through, and the direction is what the math is really about.
Side by side, what does a month actually look like?
Put the two paths next to each other for the same job and the gap is not subtle. A VA carries direct pay plus a layer of overhead you rarely price in. The tool carries a flat fee plus a small slice of your review time. Here is the monthly picture for handling payment posting across a handful of clients.
The chart only shows direct cost. Tilt it further and the VA path gets heavier, because the management hours are real hours. If your time is worth $60 an hour and you spend eight hours a month supervising, that is another $480 of value gone. The flat tool fee does not have a second column like that. It is the number, full stop.
Where does a VA still win?
A VA wins anywhere the work needs a human brain, which is most of bookkeeping. This is the part the cost math does not capture, and ignoring it would be dishonest. Only 43% of accountants have automated data entry even though 95% have adopted some automation (Intuit QuickBooks, 2025). The reason the other 57% have not is that most bookkeeping does not reduce to extraction and matching.
A VA can chase down missing documents, field a client's email, categorize the weird transactions that need context, and handle the dozen small judgment calls that fill a real workday. A tool cannot do any of that. So if your bottleneck is "I need more hands on varied work," a part-time VA is the right hire and I would not talk you out of it. The narrow comparison in this post only holds for the narrow job.
The honest line: Automation does not replace a VA. It replaces the part of a VA's day that should never have required a person, which happens to be the part that scales worst and errs most.
Where does automation win?
Automation wins on the repetitive, format-heavy translation work between a document and a ledger, and it wins decisively. AP professionals spend more than 10 hours a week processing invoices (DocuClipper, 2025, citing Medius). For a multi-client bookkeeper, a big chunk of that is reading remittances and matching them to open invoices, one at a time, in a QBO screen built for recording a single payment.
That is the exact work a tool does not get tired of. It learns the extraction logic once and applies it to every client who sends a PDF, every client who sends a CSV, every new file you connect. A VA has to be retrained for each one. When a client changes their remittance format, and they will, the tool adapts faster than a person can be re-onboarded. The messier and more variable the input, the worse a human performs against a tool built for it. Payment posting sits firmly in that category. I wrote more about that build-versus-hire decision in why I built a tool instead of hiring a VA, if you want the longer story behind this math.
How I actually run it now
I run a hybrid, and I think most practices should. The cost comparison is not an argument for firing anyone. It is an argument for putting each kind of work where it belongs. Ground Control handles the extraction, matching, and posting. I review the approvals, which takes minutes. Everything that needs a human, the client communication, the judgment calls, the cleanup, stays with a human.
The goal was never to cut people out of the practice. It was to stop using people as a data-entry bridge between two systems that should be talking to each other. Once the repetitive slice is automated, the hours you free up go toward the work clients actually pay you to think about. That is the buy-back. Not a robot doing your job, just the dullest part of it costing you under $100 a month instead of $700.
If the payment-posting grind is the part eating your week, take a look at Ground Control. It does one job, and it does it for a flat fee that does not climb with your client list.
Frequently asked questions
How much does a part-time bookkeeping VA cost per month?
US-based bookkeeping VAs commonly bill $15 to $25 an hour, so a part-time arrangement of eight to ten hours a week runs roughly $700 to $900 a month in direct pay. That figure excludes your own management time and the rework from manual errors, which run about 1.6% per invoice (ResolvePay, 2025). The true number is higher than the hourly rate suggests.
Is automation cheaper than a VA for bookkeeping?
For the narrow job of payment posting, yes, and by a wide margin. A flat tool fee under $100 a month replaces $700-plus of monthly VA labor for that specific slice, since 62% of AP processing cost is labor (APQC via ResolvePay, 2025). For general bookkeeping that needs judgment and client contact, a VA is still the better spend. The two solve different problems.
Can I use both a VA and automation?
Yes, and that is what I recommend. Let the tool handle extraction and matching, the repetitive and error-prone part. Let the VA handle document chasing, categorization, client follow-up, and review. You get better output from both because each is doing work suited to it. Only 43% of accountants have automated data entry (Intuit QuickBooks, 2025), so this combination is still an edge.
Does the automation post payments without me checking?
No. Every batch goes through an approval screen first. The tool extracts the data and proposes the QBO invoice matches, and you confirm or correct before anything posts. The judgment stays yours. The data entry is what goes away, which is the part that was never a good use of a bookkeeper's time.
The bottom line
The "hire a VA" advice is not wrong. It is just aimed at the wrong part of the problem when the problem is payment posting. For varied human work, hire the human. For the repetitive job of pulling numbers out of documents and into QBO, $700 a month of labor is paying full price for something a flat fee under $100 handles better and more consistently.
Run your own version of this math. Count the hourly rate, then count the management hours, the onboarding per client, and the rework you do not currently track. Put the flat tool fee next to it. The point is not to replace anyone. It is to stop paying labor rates for data entry, and to spend the hours you buy back on the work that actually grows your practice.
Or keep posting payments one screen at a time. I did it for years. I just stopped pretending it was free.