"I'll make it work"
He's engaged, apologetic and entirely well-intentioned. The arrears are twelve hundred, he wants it cleared, and he has done the arithmetic in his head in about four seconds.
"Four hundred a month. Three months and it's gone."
The agent's screen is happy with that. It clears the balance inside the quarter, it hits every measure on the dashboard, and the call can end in ninety seconds with everybody feeling better.
It will also fail in week six, and when it does he will be further behind than he is now, less willing to answer the phone, and convinced — with some justification — that talking to this lender made things worse.
The move that feels wrong to everyone
Talking a customer down from their own offer is counterintuitive, commercially uncomfortable and the highest-value thing an agent can do in this call. Three features make an affordable repayment arrangement conversation unlike any other in the lending cluster.
The customer is cooperative and wrong. There's no conflict to manage. He's trying to do the right thing, and the number he's produced is aspirational rather than calculated.
The failure is delayed and attributed elsewhere. The arrangement breaks in six weeks. By then it's recorded as a broken promise, and nothing connects it back to the conversation that created it.
And every incentive points at taking the number. Balance cleared, arrangement agreed, call handled. An agent who negotiates the figure downward makes today's numbers worse in exchange for an outcome that lands on someone else's report in three months.
How to spot the over-offer
It's recognisable and the signals are consistent.
The number matches the arrears too neatly. Four hundred for three months against twelve hundred owed is arithmetic performed on the debt, not on his budget.
It arrives instantly. A figure produced in four seconds hasn't been checked against anything.
"I'll make it work." The single most reliable indicator in the call. It means the sum doesn't currently work.
And nothing else is mentioned. No other creditors, no other pressures, no sense of what's left after it. If the conversation contains only this debt, the plan is being built on a fraction of the picture.
Run the arithmetic the other way
The technique is one directional change, and it's the whole thing.
Don't start from what's owed. Start from what's left. What comes in, what has to go out, what remains — and only then what portion of the remainder is realistic and sustainable.
Working from the arrears produces a number the debt requires. Working from the surplus produces a number he can pay. Most collections conversations do the first and then ask an affordability question afterwards, which is the assessment arriving too late to affect anything.
Say it out loud, early: "I'd rather agree something smaller that definitely works than something bigger that falls over in a month. Genuinely — that's better for both of us." He will not expect that sentence, and it changes what he's willing to tell you.
Ask about the others. Arrangements rarely break because of the debt being discussed; they break because of the three that weren't mentioned. "Who else is waiting on money at the moment?"
Test it against a bad month. "If the car needed something in March, would this still go out?" An arrangement that only survives a perfect month isn't an arrangement.
Leave a margin, and say why. "Let's do two hundred. If you get a good month you can always pay more — that's easy. Missing one isn't."
Signpost what exists. Where a customer's overall position is beyond this account, the firm's referral routes and free debt advice services exist for that, offered once and without pressure.
And set a check-in before the first payment is due. Not after the first failure. Before the first success.
Four ways it goes wrong
The number-taker, who accepts whatever is offered. Efficient, universally rewarded, and the direct cause of most re-breaks.
The arrears-anchored agent, who starts from the balance and negotiates downward — a process that ends at a number defined by the debt rather than the budget.
The sole-creditor assumer, who never asks what else is being paid and builds a plan on a quarter of the facts.
The no-margin planner, who arrives at a technically correct figure with nothing left over, so the first unexpected expense ends it.
Why this isn't trained
Collections training is negotiation-shaped. The model is anchoring, concession and closing at the highest sustainable number — with "sustainable" doing no real work, because nothing tests it.
Affordability is a form, and it's completed in the wrong order. In many operations the income and expenditure assessment is filled in after a figure has been agreed, which makes it documentation rather than a decision.
The consequences land outside the reporting window. Arrangements agreed in January break in March and are counted as March's problem, so no feedback ever reaches January's behaviour.
And peer role play produces a realistic offer. A colleague playing a customer names a plausible figure they've thought about. The real one names an optimistic figure instantly, out of embarrassment and goodwill — and recognising that in four seconds is the entire skill.
What sustainable arrangement training can rehearse
A simulation can hold a cooperative customer who over-offers, resists being talked down, and has commitments he doesn't volunteer — and score whether the agent worked from surplus rather than arrears and left a margin. Foretell AI supplies the counterparty configuration, transcripts and rubric-based scoring; affordability standards, forbearance options, referral routes and all regulatory requirements stay with the lender.
Four to build:
- The over-offerer, who names a figure instantly and insists he can manage it.
- The one with undisclosed commitments, where three other creditors emerge only if asked.
- The one who’s embarrassed to say a small number, testing whether the agent makes it easy to.
- The one whose surplus is genuinely nil, where no arrangement is the right answer and signposting is the outcome.
Design caution. Affordability assessment standards, forbearance options, treatment of customers in difficulty and referral obligations vary by jurisdiction and are set by the lender. Nothing here states any requirement or constitutes debt advice; modules must use the firm's own process and referral routes, and the exercise rehearses the conversation only.
Designing the module
Pass one — the recognition. Score whether the agent identified the over-offer and how quickly.
Pass two — the arithmetic. Score whether the conversation worked from income and outgoings or from the arrears figure, and whether other creditors were asked about.
Pass three — the margin. Score whether the agreed figure left headroom, whether a bad month was tested, and whether a pre-payment check-in was set.
Rubric on observable behavior: Was the customer's first figure accepted? Did the agent state a preference for a smaller sustainable amount? Was the calculation run from surplus or from arrears? Were other commitments asked about? Was a bad month tested? Was the final figure below the customer's opening offer?
"Was the final figure below the opening offer" is the measure, and it is the only rubric item in four series where a lower commercial outcome scores higher. An operation that adopts it has made a statement its agents will believe faster than any policy document.
The operator case
Arrangement-kept rate at three and six months is the number that matters, and most operations report arrangements agreed. Splitting kept-rate by agent shows immediately who is taking numbers and who is building them.
Re-breaks are expensive twice over. The lender absorbs another failed cycle and the customer's willingness to engage falls, which closes off the cheaper options later.
The sequencing is auditable today. Check whether income and expenditure assessments are completed before or after the figure is agreed. Where it's after, the assessment isn't performing the function it exists for.
And the target is the obstacle. Agents measured on arrangements agreed or value secured will take the four hundred. Measuring kept-rate, and saying explicitly that talking a figure down will never count against anyone, is the intervention — training without it is instruction to do something the scoreboard punishes.
For lending and customer operations programmes, this is a clean case of short-horizon measurement defeating a stated policy: everyone agrees arrangements should be sustainable, and the number on the wall counts the ones that were agreed.
Frequently asked questions
How do you set a repayment arrangement a customer can actually keep? Work from what's left after income and outgoings rather than from the arrears figure, ask what other creditors are waiting, test the amount against a bad month, and leave headroom deliberately.
Should you accept the amount a customer offers? Not automatically. An instant figure that exactly matches the arrears is usually aspirational, and an arrangement that breaks leaves the customer worse off and less willing to engage.
Why do payment arrangements fail? Most often because they were built on the debt rather than the budget, because other commitments weren't asked about, or because there was no margin for an ordinary unexpected expense.
What should collections teams measure on arrangements? The proportion still holding at three and six months, split by agent — not the number agreed. Counting arrangements agreed rewards taking figures that were never going to hold.
The short version
He offered four hundred because it clears the debt in three months, not because he has four hundred.
Say you'd rather agree something smaller that works. Run the sum from what's left rather than from what's owed. Ask who else is waiting. Test it against a month when the car breaks. Then agree two hundred, explain why, and check in before the first payment rather than after the first failure.
The call takes six minutes longer and produces a smaller number. It's the better outcome for both sides, and nothing on the dashboard will say so.
Foretell AI lets lenders build conversational simulations — including over-offering customers, affordability conversations and sustainable arrangement scenarios like the one above — with configurable counterparties, transcripts, recordings, and rubric-based evaluation. If you measure arrangements agreed rather than arrangements kept, we're happy to walk through how other lenders have structured it.