Pattern 01CollectionsImprove phase: in pilot

The ProcessArc Pattern Library

Why your collections team is losing hours to the wrong work.

A pattern we see across banks and credit unions. One institution lived with it for more than three years. We diagnosed and redesigned it in four months.

Get the full pattern (free PDF)Score your operation (2 minutes)

Or calculate what this pattern costs you

Chapter 1 of 7 · The pattern

A commodity industry has commodity problems.

Financial services is a commodity business wearing a thousand different logos. It's heavily regulated, the products are largely interchangeable, and the back-office processes behind them look remarkably alike.

When one institution has a process problem, the odds are good that a dozen others have the same one, wearing a different name.

Every delinquent account costs more to work than a current one. When the people paid to work those accounts spend their day on manual admin instead, the cost compounds.

This pattern walks through one real engagement. Not because it's unusual, but because it isn't.

Chapter 2 of 7 · How to spot it

6 signs this pattern is in your operation.

If three or more sound familiar, the rest of this pattern is probably about you.

01
Collectors doing data work
Collectors spend an hour or more a day on manual tasks that have nothing to do with calling.
02
Too much time on exceptions
More than 15% of the team’s time goes to manual review or exception handling.
03
Decisions wait for a report
Suspensions, transfers and notices are triggered by a daily or weekly report, not live account status.
04
Exclusions slip through
Bankruptcy, hardship, protected funds or closed accounts get missed in manual review.
05
Same task, different answers
Several teams take the same account action, with no single source of truth.
06
The backlog never clears
Work rolls to the next business day because the team can't clear the day's volume.

Next step · Self-assessment

Does this pattern sound like your operation?

8 yes/no questions, scored out of 9. It isn't a scientific instrument: it's a fast way to see whether this pattern applies to you. Your score is free.

0–2: Ahead of the pattern
3–5: Moderate exposure
6–9: Strong match with this pattern

Question 1 of 8Pattern 01: Collections

Has your collections process gone more than 12 months without being mapped end to end, step by step?

Counts double. An unmapped process is where the other risks hide.

Takes about two minutes

Next step · ROI calculator

Put your own number on it.

Think of it as the fully-loaded cost of a phantom employee: someone who shows up every day but never picks up the phone. Enter your own numbers, then register to see your result.

Staff who lose part of their daye.g. collectors pulled off the phoneHours each of them loses a dayhrsStaff full-time on a manual taske.g. money transfers, manual review

Fully-loaded hourly cost per FTE$Salary plus benefits and overheadTarget cut in off-phone time%Working hours a dayhrsWorking days a yeardays

Pre-filled with an example that reproduces this pattern's totals: 7 FTE equivalents, 14,000 hours and $880K a year. Replace the numbers with your own.

Your numbers

Capacity lost: •.• FTEs
Hours lost a year: ••,•••
Cost a year: $•••,•••

See your numbers, and get the full pattern with them.

Name and work email. Your numbers stay with ProcessArc.

Chapter 3 of 7 · What it costs

14,000 hours a year on work that wasn't collecting.

Collections' core job is outbound calling. A growing share of the team's day was going somewhere else.

~2 hrs a dayof collectors' time went to manual “special tasks” instead of calls
4 FTEsworked 100% on processing money transfers
7 FTEsin total: full-time equivalents spent on work that wasn't collecting

14,000

hours a year

of collector capacity spent on work that wasn't collecting

$880K

a year in lost capacity

never budgeted: it showed up as a team that was always a little behind

How we got the number

7 FTEs × 8 hrs × 250 days = 14,000 hrs × $62.80 = $879,200

$62.80 is the fully-loaded hourly cost of one full-time equivalent (FTE). The hours came from measuring what the team actually did, not from estimates.

Chapter 4 of 7 · Why it happens

3 processes, one root cause.

Not a technology conversation: we mapped the process, walking the actual steps staff performed, in order.

100% manual, 100% repeatableEvery step was done by hand, yet followed the same logic every time.
The reports were a bottleneckIncoming reports needed manual clean-up before anyone could act.
Never enough hoursThe day's volume never cleared. Work rolled over, every day.

Right to Cure letters

Manual review, extraction, filtering, validation and letter generation
Five human touch points for a decision with no real judgment call in it

Money transfers

Closed or ineligible accounts pulled in by error
Exclusions missed: revoked cards, bankruptcy, protected funds
Business cards misclassified; prior-day duplicates not filtered

Line of credit suspensions

Several teams suspending credit lines inconsistently
Decisions based on periodic reports, not live delinquency status
Accounts crossed the 45-day line before anyone saw it

The root cause.These decisions needed to happen in real time, the moment an account became eligible. Instead they waited for a periodic report and a person to catch them. The data was there and the logic was repeatable: it could have been automated all along.

Chapter 5 of 7 · What fixes it

What to automate, and what to keep human.

Most automation vendors skip this part, because their business model depends on you automating everything. Ours doesn't.

Automate

Rules-based filtering, exclusion and validation logic
Report clean-up and standardization before human review
Standard letter generation, such as Right to Cure letters
Real-time eligibility triggers in place of periodic batch reports

Keep human

Hardship and exception cases that need member-specific judgment
Ambiguous or flagged accounts, routed for compliance sign-off
Final review of anything touching regulated communication timing, frequency or channel

Guiding principle

Collections communication is heavily regulated on frequency, timing and channel.

Automation doesn't remove those guardrails. It has to be built inside them, with mandated human review for anything high-risk or ambiguous.

Chapter 6 of 7 · Where it stands

Target: ~$440K a year.

Recognizing the pattern on contact, instead of starting from zero, is what closed the gap.

3+ years

living with the bottleneck: unmeasured, treated as just how the work was

4 months

from first process map to a redesigned, tested solution, with ProcessArc

~$440K a year

Cutting off-phone time in half recovers about half of the $880K lost each year, returned as collector capacity without adding headcount.

Project status

Define, Measure, Analyze, Improve, Control (DMAIC): the five phases of a Lean Six Sigma project.

Define, Measure, AnalyzeComplete
ImproveIn pilot
ControlNext

The automation logic is running in a controlled pilot. We will add measured results to this pattern when the pilot completes. We show the work while it is still being proven, because finished, polished outcomes teach you less.

Chapter 7 of 7 · The pattern, restated

This wasn't unique to one institution, one core system, or one collections team.

This happens when repeatable, rules-based work sits inside a manual process long enough that nobody questions it, and when decisions that should run continuously run instead on a report pulled once a day or once a week. Recognizing it took a conversation. Fixing it took four months.

Next step · Structured thinking template

The 5 questions we ask, in order, every time.

You don't need a Black Belt to start. You need honest answers.

01DefineWhat's the process, who owns it, and what's the goal in plain numbers (not “improve efficiency”)?
02MeasureWhat does the process actually do today, step by step, not what people think it does?
03AnalyzeWhere does the manual work break: missed exclusions, stale reports, inconsistent handling across teams?
04ImproveFor each broken step: is the logic repeatable (automate) or judgment-based (keep human)?
05ControlHow will you know the fix is holding six months from now: what's the metric, and who owns it?

Take it with you

Pattern 01: Collections (PDF, 10 pages)

The full pattern, ready to share with your team.

More patterns: see the Pattern Library

Free · Takes 20 seconds

Read the rest of Pattern 01: Collections.

One short form unlocks everything on this page:

Chapters 3 to 7: what it costs, the root cause, what to automate and what to keep human, and where it stands
Your own result from the ROI calculator
The structured thinking template we use on every engagement
Pattern 01: Collections (PDF, 10 pages), to share with your team

Every pattern starts with a conversation.

Tell us about the process that keeps your team behind. The first consultation is free.

Discuss a business problem See our case studies

Why your collections team is losing hours to the wrong work.

A commodity industry has commodity problems.

6 signs this pattern is in your operation.

Collectors doing data work

Too much time on exceptions

Decisions wait for a report

Exclusions slip through

Same task, different answers

The backlog never clears

Does this pattern sound like your operation?

Put your own number on it.

Read the rest of Pattern 01: Collections.

Every pattern starts with a conversation.