An accounting firm recovered $67K in 5 weeks by automating AR follow-ups. Here's exactly how we built it.
Article text
The office manager had three monitors. One displayed QuickBooks. One had a spreadsheet with 340 rows of client invoices, color-coded by status: green for paid, yellow for 30+ days, red for 60+. The third monitor showed her email, where she was manually copying invoice numbers into follow-up templates she'd written herself.
She'd been doing this every Tuesday and Thursday for four years.
When we asked how many invoices were currently overdue, she said, "Let me check." She scrolled for 90 seconds. "Maybe 40? No, wait." More scrolling. "Probably closer to 55."
It was 73. We counted.
That gap between "probably 55" and the actual 73 was costing the firm roughly $67,000 in receivables that nobody was chasing consistently. Not because clients wouldn't pay. Because nobody was asking.
If your accounting firm is chasing invoices manually and suspects money is slipping through the cracks, we do free 30-minute discovery calls to map exactly where the leaks are.
The Problem Nobody Wanted to Quantify
This was a 22-person firm in the Midwest. Tax prep, bookkeeping, advisory work. Solid clients, mostly small businesses and high-net-worth individuals. Revenue around $3.2M. They weren't struggling. They were just leaving money on the table.
Their invoicing process worked fine. Invoices went out on time through QuickBooks. The problem started the moment an invoice wasn't paid within 30 days.
Here's what the follow-up process looked like:
Office manager opens the master spreadsheet (updated manually from QuickBooks)
Sorts by date to find overdue invoices
Checks each client's payment history in QuickBooks to see if there's a pattern
Drafts a follow-up email using one of three templates (15 days, 30 days, 60+ days)
Copies the invoice number, amount, and due date into the email
Sends the email and logs that she sent it in a separate column on the spreadsheet
If no response in two weeks, flags it for the partner to call
Seven steps. Every invoice. Twice a week.
The office manager estimated she spent about 6 hours a week on this. We timed it over two weeks. It was closer to 9.
And even with 9 hours a week, invoices were falling through. Some clients got their first follow-up at 45 days instead of 15. Some never got followed up at all because the spreadsheet row was accidentally hidden. One invoice for $8,400 had been sitting unpaid for 114 days. Nobody noticed because the row was filtered out of the default view.
What We Found During the Diagnosis
We spent the first week doing what we always do: mapping the process before touching any technology. We sat with the office manager, watched her work, and asked questions.
Three things jumped out immediately.
The data lived in two places that didn't talk to each other.
QuickBooks had the real-time invoice status. The spreadsheet had the follow-up history. Every time the office manager wanted to know "has this client been contacted about this invoice?", she had to cross-reference two systems manually. That's where errors crept in.
The follow-up cadence was inconsistent.
In theory, clients got a reminder at 15 days, 30 days, and 60 days. In practice, the timing depended on which Tuesday the office manager happened to check. Some weeks she was buried in tax season work and skipped a session entirely. During March and April, follow-ups basically stopped.
Nobody had visibility into the full AR picture.
The partners knew AR was "fine" because the firm was profitable. They didn't know the actual DSO (days sales outstanding) was 52 days. For a firm their size, that number should be closer to 25-30. Every extra day of DSO is cash sitting in someone else's account instead of yours.
We pulled every invoice from the previous 12 months and ran the numbers. The firm had written off $23,000 in the past year as "uncollectable." When we dug in, $14,000 of that was from clients who simply never got a follow-up after the first reminder. They would have paid. Nobody asked.
What We Built
The solution wasn't complicated. That's the point. The office manager's process was solid in concept. The execution just couldn't keep up because a human was doing work that a workflow handles better.
We built an automated follow-up system that pulled directly from QuickBooks and ran on a simple set of rules:
Day 15 past due:
Polite reminder email. "Just a friendly note that invoice #[number] for $[amount] was due on [date]. Let us know if you have any questions." Personalized with the client's name and the specific partner they work with.
Day 30 past due:
Firmer follow-up. Different tone, same professionalism. Includes a direct payment link. "We want to make sure this doesn't get lost in the shuffle."
Day 45 past due:
Flagged for the assigned partner with a summary: client name, amount, days overdue, and a log of every previous communication. The partner gets a Slack notification and an email. One click to review the full history before picking up the phone.
Every step gets logged automatically. No spreadsheet. No manual data entry. No hidden rows.
The whole thing checks QuickBooks every morning at 6 AM, identifies any invoice that crossed a threshold overnight, and fires the appropriate action. The office manager gets a daily digest at 8 AM showing what went out, what got flagged, and what's coming up.
Build time: 3 weeks. Two weeks for the core workflow and QuickBooks integration. One week for testing with live data (we ran it in parallel with the manual process to make sure nothing fell through).
Monthly running cost: under $80.
The Results
We turned the system on and let it run for 5 weeks before measuring.
$67,000 recovered.
That's not projected savings. That's cash collected from invoices that were sitting untouched. Some had been overdue for 90+ days. The automated follow-ups restarted conversations that had gone cold.
DSO dropped from 52 days to 29 days.
That's 23 days of cash flow the firm got back. For a $3.2M firm, that translates to roughly $200K in improved working capital position at any given time. According to recent AR benchmarking data , firms with automated AR processes average 40 days DSO compared to 47 for manual processes. This firm beat the automated average by 11 days because the follow-up cadence was tighter than most off-the-shelf tools allow.
The office manager got 9 hours a week back.
She didn't get reassigned. She started spending that time on client onboarding and billing accuracy work that had been perpetually deprioritized. Within two months, billing errors (wrong rates, missed time entries) dropped by about 40% because someone finally had time to review them before invoices went out.
Write-offs dropped.
In the 3 months after launch, the firm wrote off $1,200 in uncollectable invoices. The previous 3-month average was $5,750. That's a 79% reduction.
The follow-up workflow went from 7 manual steps to 2: the system handles everything automatically, and the partner steps in only when an account hits the 45-day flag. That's it.
Why This Worked (And What Almost Didn't)
The biggest risk wasn't technical. It was tone.
Automated follow-up emails from accounting firms can feel impersonal and aggressive if you're not careful. The managing partner was worried that clients would feel like they were being hounded by a robot. Fair concern.
We spent a full day on the email copy. The 15-day reminder reads like it came from a real person because it references the specific engagement (not just "your invoice"). The 30-day email includes the payment link but also offers to hop on a call if there's a billing question. The 45-day partner flag doesn't trigger an automated email at all. It's a handoff to a human.
Two clients actually replied to the 15-day reminder saying, "Oh sorry, this totally slipped. Paying now." They thought they were replying to the office manager. That's the bar for good automation: the recipient shouldn't be able to tell.
The other thing that almost tripped us up was the QuickBooks sync. The firm was on QuickBooks Online, which has a solid API. But their chart of accounts had some custom categories that made invoice-status mapping tricky. We spent about 3 days on data mapping alone. If they'd been on desktop QuickBooks, the integration would have taken twice as long. This is the kind of thing you can't predict until you're inside the system, which is why we always start with a process diagnosis before writing any code .
The Bigger Picture
61% of finance leaders still rely on spreadsheets and manual processes to manage AR . That stat is from 2025, and it tracks with what we see in the field. Most accounting firms we talk to know their AR process is inefficient. They just haven't quantified how much it's costing them.
The math on this project was straightforward: $67K recovered in 5 weeks, against a build cost of roughly $4,500 and a monthly run rate of $80. That's a payback period measured in days, not months.
But the real value wasn't the recovered cash. It was the visibility. For the first time, the partners could see their full AR picture in real time. No more "probably 55" when the answer was 73. No more end-of-year surprises when the write-off column was bigger than expected.
If your firm is chasing invoices with spreadsheets and email templates, you're almost certainly leaving money on the table. The question isn't whether automation would help. It's how much you're currently losing by not having it.
Send us a message and we'll map your AR process for free. Even if you don't build anything with us, you'll walk away knowing exactly where the leaks are.