Back to blog

How a 12-person agency recovered €47,000 in overdue invoices in 60 days

Published on: May 2, 2026

min read

Agency Recovery Case Study
case study AR automation agency

This case study is composite, not a real named client. Every number, sequence, and outcome is built from documented case studies of small firms that adopted structured AR workflows. If you run a 5 to 30-person service agency, the mechanics here mirror what an AR Recovery System installation actually looks like.

Why this case study exists

This is a composite scenario. It is not a real named client. The numbers, sequence, and outcomes are built from documented industry benchmarks and published case studies of small firms that adopted structured AR workflows. Sources for every range and figure are cited at the end of this post.

If you run a 5 to 30-person B2B service agency, IT services firm, or consultancy, the structure here mirrors a typical AR Recovery System rollout based on documented industry case studies.

The starting point: what the agency looked like before

A 12-person creative agency in Western Europe. Monthly billing around €120,000. Mix of monthly retainers and project work. Standard payment terms: 30 days net. No dedicated AR clerk. Founder handles finance “on Fridays”.

What their AR situation looked like

At baseline, the agency had €68,000 in invoices past due, spread across 19 different clients. The longest-overdue invoice was 87 days late. This is consistent with documented patterns: a 12-person consulting firm profiled by Builts.ai in 2026 carried over $200,000 in receivables on $1.8m annual revenue — more than a full month of billing locked up. For a 12-person agency, having 30 to 100% of one month’s revenue in overdue AR is realistic without structured follow-up.

The founder spent 6 hours per week on follow-up emails, calls, and reminders. Two of those hours were on weekends. This sits within the range Xero research and Upflow report for small business owners: roughly 10% of working time spent chasing late payments. Zendu’s 2025 analysis of manual follow-ups puts the upper end at 14 hours per week for the worst-affected SMBs, so 6 hours is mid-range, not extreme.

Why nothing was working

Three real problems behind the numbers.

First, reminders went out inconsistently. Some clients were chased after 3 days, others were forgotten for 6 weeks. There was no system, just whoever had time.

Second, the same template went to everyone. Long-term retainer clients got the same firm reminder as a new project client three weeks late. Reliable clients felt insulted; chronic late payers ignored the firm tone because there was no escalation behind it.

Third, the founder hated chasing money, so he procrastinated. Reminders that should have gone day 1 past due went day 14. By then the client had moved on to the next priority.

What was installed

An AR Recovery System wired into their existing Xero account, Gmail, and Slack. No new software for the team to learn. No portal for clients to log in to.

How the system worked

The core flow: every morning, the system pulled the AR aging report from Xero. For each overdue invoice, it applied a cadence based on three variables: how late the invoice was, the client’s historical payment behaviour, and the invoice amount.

The cadence followed the sequence pattern documented in the Builts.ai case study and adapted by similar AR specialists: pre-due reminder around day 25, first overdue reminder day 32, follow-up day 39, firmer follow-up day 46, escalation day 53.

For reliable clients 1 to 7 days late: a friendly nudge, signed by the founder, copying the project manager. Tone matched their existing relationship.

For reliable clients 8 to 21 days late: a firmer reminder noting the original due date and offering a payment link. Bluevine’s 2026 Payment Gap Report data shows invoices with embedded “pay now” buttons are paid 174% faster on average (7 days vs 18 days), so the link is not cosmetic — it materially shortens time to cash.

For chronic late payers (clients with 3 or more late payments in the prior 12 months): tighter cadence starting day 1, escalation to a payment-plan offer at day 14, and a Slack notification to the founder at day 21 to make a personal call.

Disputes (any reply containing words like “issue”, “wrong”, “discuss”, “hold on”) were routed to a human inbox immediately, never auto-replied.

What stayed manual

Decisions involving relationships. Writing off bad debt. Negotiating payment plans for amounts over €5,000. Deciding when to pause work for a non-paying client.

The system handled the predictable 80%. The founder kept the 20% that needed judgement.

The 60-day results

Recovery odds depend heavily on invoice age. Yonovo’s 2026 timing synthesis of collection-agency and AR data gives the curve: 1-30 days overdue recovers at 90-98%, 31-60 days at 75-85%, 61-90 days at 50-70%, 91-120 days at 30-50%, 121-180 days at 15-30%. Most of the agency’s €68,000 backlog was under 90 days, which made aggressive 60-day work productive.

At day 30:

  • 11 of 19 overdue invoices fully paid (€26,400 collected)
  • 4 invoices on agreed payment plans (€9,200, paying 25% per week)
  • 5 invoices acknowledged with a confirmed pay date
  • 2 invoices in dispute, escalated to founder

At day 60:

  • 17 of 19 original overdue invoices closed (€38,800 fully paid, €8,200 in active plans)
  • DSO on new invoices dropped by approximately 9 days (from a baseline of around 41 days to 32 days). This sits within Transformance AI’s documented range of 8 to 15 days of DSO improvement in the first 90 days of an AI-driven collections process.
  • Founder time on AR dropped from 6 hours per week to roughly 45 minutes (mostly reviewing the dispute queue). The Builts.ai case study reports an 85% reduction in weekly AR follow-up time for a comparable 12-person firm.
  • 1 chronic late payer was offboarded after refusing a payment plan; that decision happened in week 5

Total recovered or actively being recovered in 60 days: €47,000 across collected invoices and active payment plans, plus the productivity recovery of roughly 22 hours per month of founder time.

The numbers behind the math

Setup cost of an equivalent AR Recovery System: typically €1.5k–2.5k one-time per industry estimates, with no recurring fee.

Return at day 60: €47,000 in cash flow unlocked plus founder time worth approximately €3,300 per month at €150 effective hourly rate.

ROI: setup paid back inside the first 4 working days.

What didn’t happen

No client churned because of “too many reminders”. Every cadence was tested for tone, and the system stopped sending the moment payment was confirmed (Xero webhook trigger).

No legal action was needed. The escalation path topped out at a personal founder call before going to a collections agency, and no invoice in this batch reached that point.

What surfaced unexpectedly

When reminders became consistent, hidden disputes appeared. This is well-documented: IOFM benchmarks place dispute rates at 3-5% of all invoices, but UK BEIS data summarised by Equisettle shows dispute-related delays touch 40% of B2B transactions and add 23 days to payment cycles. The agency’s 2 surfaced disputes (about 4% of active open invoices) sat right in IOFM’s range.

Two clients had legitimate billing concerns nobody had flagged because nobody was reading the replies. Resolving them quickly turned both into more loyal clients.

What the founder said worked best

Three things, in his words.

First, the system removed the emotional load of chasing money. He stopped dreading Mondays.

Second, the dispute routing was the unexpected win. Surfacing those silent issues was worth more than the cash recovered.

Third, the data became a planning tool. After 60 days he could see which clients were structurally late versus occasionally late, and adjust contracts at renewal. Industry commentary suggests roughly 20% of clients in any small B2B firm are chronic late payers requiring different treatment, though there is no robust quantitative benchmark for what percentage of clients firms actually offboard.

How to apply this if you run a similar agency

Three steps to take this week.

Pull your AR aging report from Xero or QuickBooks. Note the total overdue and the count of invoices over 30 days late. That number is your unrealized cash. The deeper context lives in our hidden cost analysis.

Count the hours your team (and you) spend per week on AR follow-up. Multiply by your effective hourly rate.

If the sum of recovered cash potential (typically 60 to 85% of overdue under 60 days) plus annualized time savings is more than 4x the cost of an AR system, the case for installing one is already made. For broader market context on why this matters now, see our 2026 cash flow statistics roundup.

The alternative isn’t “keep doing it manually”. The alternative is keep losing 10 to 20% of cash that’s already yours, plus the founder time that should be on growth.

Sources

  • Builts.ai 2026 case study of 12-person consulting firm AR transformation
  • Yonovo 2026 invoice age recovery timing synthesis
  • Upflow 2024 AR statistics report
  • Xero late payments research summarised in UK media
  • Zendu 2025 “Hidden Cost of Manual Invoice Follow-ups”
  • Transformance AI DSO reduction benchmarks 2024-2026
  • IOFM (Institute of Finance and Management) dispute rate benchmarks
  • UK BEIS dispute data summarised by Equisettle
  • Bluevine Payment Gap Report 2026
  • Atradius Payment Practices Barometer Western Europe 2024

FAQ

Is recovering €47,000 in 60 days realistic for a 12-person agency?

Yes, within documented industry ranges. Case studies of small firms adopting structured AR workflows show overdue balances dropping by 60 to 67% over a few months. Recovery odds are highest for younger invoices: 90 to 98% at 1-30 days overdue, 75 to 85% at 31-60 days, falling to 50-70% at 61-90 days according to Yonovo’s 2026 synthesis. A €47,000 recovery from a €60-80k backlog under 90 days fits this curve.

How quickly does DSO move after installing an AR Recovery System?

Industry data points to 8 to 15 days of DSO reduction within the first 90 days of an AI-driven collections process (Transformance AI 2024-2026), with longer-term improvements of 10 to 30 days as processes mature (ProcIndex 2026 AR Automation Guide). Most early gains come from consistent pre-due reminders and removing invoicing errors, not aggressive escalation.

How much founder time can AR automation realistically save?

Survey data from Upflow and Xero suggests small business owners commonly spend around 10% of their working time chasing invoices. Zendu’s 2025 analysis puts the upper bound at 14 hours per week for the worst-affected SMBs. The Builts.ai 2026 case study of a 12-person firm reports weekly AR time dropping from 8+ hours to under 1 hour after automation, an 85% reduction.