Menu

GGA RESEARCH & METHOD

Most of Your Overdue Book May Be Waiting on Your Own Organisation

Last reviewed: 9 November 2026

Fifteen Indian receivables engagements, across nine sectors and several million invoice lines, examined for a single question: when an invoice is unpaid, what is actually preventing payment? The answer is consistent enough to be uncomfortable. In engagement after engagement, the majority of what a company reports as overdue is not money a customer is refusing to pay. It is money the company has not yet made payable.

38.5%Of one opening book blocked before a customer was contacted
78.0%Of blocked value owned internally on a logistics book
61.4%Owned internally on a maintenance book
481 of 481Accounts where assignment timing was decided by the wrong function
2.4%What a receivable returns once nobody has acted for five years

The argument

A receivable does not go unpaid because a customer decided not to pay. It goes unpaid because somebody inside the seller did not do something — and nobody is measured on whether they did.

01

Executive summary

How to read the evidence in this paper. measured computed directly from client line-item data. modelled arithmetic built on stated assumptions. observed a pattern recorded across assignments rather than calculated. Every percentage states the denominator it is taken against; figures on different denominators are not comparable and are never averaged.

This paper does not describe a single engagement. It sets out what fifteen engagements have in common, because the pattern is more useful to a CEO than any individual result.

  1. Blocked receivables constrain growth and consume the most expensive time in the business. Capital that has not returned cannot be redeployed, and the chasing falls on salespeople and field engineers — the people least able to spare the time and least suited to the task. The people doing the obstructing almost always believe they are protecting the company, which is why the correction cannot come from among them.
  2. A large share of overdue value could not be solved by asking the customer to pay. On a logistics book, 78.0% of blocked value required action by the seller's own commercial, billing, pricing or operations functions. On a maintenance book, 61.4%. On a technology services book, 38.5% of the entire opening balance could not be collected by anybody until the seller fixed it first.
  3. The blockers are the same everywhere. Agreements never signed, purchase orders never issued, invoices that cannot be delivered, deliveries never evidenced, customer onboarding never completed, credit notes never passed, approvals sitting on a desk. Nine sectors, one list.
  4. Information is withheld rather than shared. On one engagement the foreign parent could see the balance and nobody internally could explain it. On another, one function controlled both what data was released and which accounts went for recovery.
  5. Onboarding is sat on, and it returns. On a testing book, incomplete customer KYC held 13.8% of collectable value, was cleared to 0.3% over nine months, and rebuilt to 9.7% — because every new customer arrives without it and nobody upstream owns that.
  6. Unclosed issues age into the buckets where money dies. Recovery yield on a trade channel book fell from 35.2% inside a year to 8.1% beyond two and 2.4% beyond five. Every month an internal issue stays open, the balance behind it moves toward the part of the curve where nothing comes back.
  7. The partner brought in to help is read as a competitor. An external team is complementary capacity, but the person who has managed the book experiences it as an assessment of their own performance — and behaves accordingly. The behaviour intensifies with tenure, because the longer somebody has held a book the more of its aged balance accumulated on their watch.
  8. The information reaching the board is filtered before finance ever sees it. A finance team can only test what it is given, and what it is given is the account owner's version of each balance. Without the underlying ledger detail it cannot verify that account, however much it wants to — and a receivable outlasts the tenure of almost everyone who touches it, so nobody ever sees a full cycle from sale to write-off.
  9. The internal team frequently reads an outside partner as competition, and the response is to withhold. Data arrives late or partial, accounts are nominated selectively, and on one recently begun engagement contractual terms were extended from 45 days toward 90 without any change in what customers actually paid — which reclassifies overdue balances as current without collecting anything.
  10. Protecting the receivable is a capital decision, not an administrative one. Money not collected inside its window is destroyed rather than delayed — and the capital destroyed belongs to the shareholder. Framed as an admission of failure, the conversation is resisted; framed as capital protection, it belongs at board level where the authority to act actually sits.
  11. And the decision to act belongs to the wrong function. On a 481-account assignment tested in full, the value released for recovery matched the already-aged balance on every single account. The organisation kept what was young and released what was already lost.
02

What blocked receivables actually cost

Before the evidence, it is worth being precise about what is at stake — because a blocked receivable is usually discussed as a finance inconvenience, and it is not one.

It constrains growth directly

Capital that has not returned cannot be deployed again. On a leasing book collecting roughly nine rupees in ten within the month, that conversion rate sets how many times the same capital turns in a year — and therefore how fast the business can grow. On a trade channel book where balances aged past five years returned 2.4%, the capital did not come back at all.

A company with a slow book grows at the speed of its collections whether it knows it or not. Sales targets, capacity plans and investment decisions are all constrained by the same figure, and it is rarely the figure being managed.

It consumes the most expensive time in the business

An unresolved receivable does not sit quietly. It generates a monthly review, an escalation, a customer call, a commercial discussion, a provisioning debate. Multiply that across a few hundred accounts and the cost is measured in senior attention rather than in rupees.

And it falls on the people least able to spare it. On one engagement the client's field engineers were being pulled into chasing maintenance renewals — 3,868 of which were worth under ₹25,000 and carried 2.3% of the value. On several others, commercial engineers and salespeople were spending their weeks resolving paperwork rather than selling. The most expensive people in the business, applied to the least valuable work available.

Every hour a salesperson spends chasing an invoice is an hour not spent selling, and the invoice is chased badly — because the person doing it is measured on the next order, not the last one.

And the people doing it believe they are helping

This is the part worth stating carefully, because it is the reason the problem is durable. A salesperson protecting a customer relationship from a collections conversation is acting to preserve future revenue. A credit controller who does not escalate is protecting a working relationship they will need again next month. A finance director who accepts the sales account of a balance is keeping a functioning management team functioning.

Every one of those is a defensible judgement in isolation, and together they produce a book where half the value is waiting on the company's own actions. Nobody in the sequence intends the outcome and nobody individually causes it.

The people obstructing the process almost always believe they are protecting the company. That is precisely why the correction cannot come from among them.

What a write-off actually costs

A balance written off is not a one-for-one loss. It is a loss of profit, and profit has to be earned again at the margin the business actually makes.

Replacing ₹1 crore written off, after three years of carrying it
At 8% EBITAt 10% EBITAt 20% EBIT
Additional sales needed to replace the principal₹12.50 cr₹10.00 cr₹5.00 cr
Financing already spent carrying it for three years₹0.30 cr₹0.30 cr₹0.30 cr
Total economic loss₹1.30 cr₹1.30 cr₹1.30 cr
Additional sales needed to restore the position₹16.25 cr₹13.00 cr₹6.50 cr
Even at a healthy 20% EBIT, a single crore lost takes ₹6.5 crore of new business to undo — and the business will need to find it.

And the replacement sale is not free either. ₹6.5 crore of new business at 90 days ties up ₹1.60 crore of working capital and costs a further ₹0.16 crore a year to finance. The company runs harder to stand still.

Two costs compound quietly behind that. A lengthening receivable raises working capital borrowing, which affects the rate on all borrowing rather than just the incremental facility — and lenders and rating agencies read debtor days as a governance signal, not merely a liquidity one. A book that ages visibly is priced accordingly across the balance sheet.

03

The same finding, nine sectors

Each engagement below was analysed independently, from the client's own line-item data, with no expectation of a common result. The consistency is what makes the finding usable.

What was blocking payment, by engagement
SectorWhat was blockedScaleWhose side
Technology servicesOrder-to-Invoice defects — no agreement, no purchase order, undeliverable invoice, unproven usage38.5% of the opening bookSeller
Logistics and freightCommercial, billing and pricing decisions, plus operations documents — proof of delivery, trip certificates78.0% of blocked valueSeller
Equipment maintenanceCompliance and documentation, deviation approvals, billing corrections, cancellations61.4% of blocked valueSeller
Testing and certificationCommercial intervention, deliverable documents, customer KYC incomplete27% of collectable value, on two separate booksSeller and customer admin
Process automationLive dispute pool requiring the business unit or regional head rather than collections40.5% of the dispute poolSeller, outside collections
Government medical supplyReceipt certificates uncollected, portal submissions never lodged, tranche-level objection unreconciled₹1.90 cr held for seven months against a ₹11.19 lakh actual issueSeller
Project and EPCRunning account bill not reconciled to tax invoice; deductions applied without breakdownMajority of recovery came from reconciliation, not persuasionSeller and customer
LeasingVendor registration and onboarding documents pendingTracked as a standing categoryCustomer admin, obtainable by seller
Electronic security systemsSix blocker families; only one responds to a payment reminderFive of six families held the majority of the open bookSeller and customer
Share of blocked value requiring action by the seller, by engagementBar chart. Logistics 78 per cent. Equipment maintenance 61.4 per cent. Process automation dispute pool 40.5 per cent. Technology services opening book 38.5 per cent. Testing and certification 27 per cent. Share of blocked value requiring action by the seller, by engagement Logistics 78.0% Equipment maintenance 61.4% Process automation 40.5% Technology services 38.5% Testing & certification 27.0%
Nine sectors, fifteen engagements. In every case examined, the portion of the overdue book that was genuinely a customer declining to pay was the smaller part of it. Note the denominators differ — 38.5% is of an opening book, 78.0% and 61.4% are of blocked value, 27% is of collectable value and 40.5% is of a dispute pool. They are not a comparable series and are not averaged.

This is not an argument that customers always pay. They do not. On a trade channel book, ₹10.54 crore sat with dealers who acknowledged the balance and refused. What the evidence shows is that the refusing customer is a minority of the problem, and that treating the whole book as though it were that customer misdirects almost all of the effort.

04

How the organisation blocks it

Four recognisable behaviours, then one governance failure that outweighs all of them. None is malicious. All are invisible on an ageing report.

The four internal blockers
BehaviourWhat it looks likeEvidence
The invoice is never made payableNo signed agreement, no purchase order, no deliverable invoice, no proof of delivery. The invoice is not late — it never entered the customer's payment queue.38.5% of one opening book; ₹30.46 crore of invoices cancelled or revised on another
Information is not sharedData arrives late, partial, or with a commentary that pre-empts the analysis. Questions are met with reassurance rather than records.One client's parent could see the balance and nobody internally could explain it
Onboarding is sat onCustomer KYC and vendor registration left incomplete, so a valid, undisputed invoice is unpayable by the customer's own systems.One book: onboarding blockage cleared from 13.8% to 0.3%, then rebuilt to 9.7%
Open issues are never closedThe balance behind an unresolved issue ages — and in a channel book, ageing removes the counterparty, not just the willingness.Yield fell from 35.2% inside a year to 2.4% beyond five; ₹5.81 crore owed by dealers who could not be located

The onboarding example proves these are processes, not backlogs. Nine months of sustained work cleared the category almost entirely; twelve months later it stood where it started, because every new customer arrives without a record and nobody upstream owns that link. A strong sales month produces a larger blocked balance sixty days later.

And the timing decision sits in the wrong place

On one engagement an assignment tranche was tested line by line. 481 dealer accounts prepared for release to recovery. On 481 of 481, to the rupee, the value released equalled the balance already beyond 240 days. Everything younger was retained.

The rule releases nothing before 240 days — but the pool released was not merely eight months old. On that book 41.4% of the balance was already beyond two years, where yield is 8.1% against the 35.2% available inside the first year. Recovery received the tail rather than the threshold, and the outcome was determined before any collection work began.

The formal rule was age. The discretion was over which accounts reached the list — and that sat with the function whose incentive runs the other way.

The motive is not mysterious. If an outside team collects a balance the internal team has called difficult for two years, that is a judgement on somebody. Nominating accounts already beyond hope moves difficult conversations off the desk, and poor recovery afterwards confirms what was said about using an outside firm. The test that would settle it — comparing recovery on retained balances against released ones, matched for age — is never run.

05

Why it persists

Four mechanisms that keep internal blockers invisible
MechanismHow it works
Credit control is measured on the wrong thingThe target is set on amount collected, never on which ageing bucket the money came from. A team hitting that target has done exactly what was asked of it — the measure, not the effort, is what leaves the aged book untouched.
Receipt application conceals the disputeMost ERPs apply a receipt to the oldest invoice first, so a disputed invoice is settled by money meant for current purchases. The dispute never ages, never appears overdue, never escalates.
Nobody wants to reverse revenueReversal restates a reported number, and the person who booked it has often moved on. One book carried ₹3.07 crore identified as uncollectable in month one and still carried in month eleven.
Terms are quietly extendedOn one engagement terms moved from 45 days toward 90 with no change in what customers actually paid. Extending terms collects nothing — it reclassifies an overdue balance as a current one.
A book where terms have been extended to match how late customers already pay is not a book that has improved. It is a book where the measurement has been moved to meet the result.

The cost compounds in three directions. Cash is delayed. Tax is paid on revenue recognised and never received — income tax always, and on taxed supply lines output GST is payable at time of supply with no general mechanism to recover it on a failed receivable. And the receivable reported to the board describes an asset that partly does not exist.

Nobody suffers from a carried bad receivable except the company and its shareholders. The customer keeps the goods. The person who booked the sale keeps the credit. The tax has been paid.

06

Why the correction must come from leadership

An outside partner is a complement to a credit function, not a replacement — and is very often read as competition anyway. The response is quiet rather than hostile: data arrives late, accounts are nominated selectively, questions are met with reassurance.

It correlates with tenure. The longer somebody has been in a role, the more of the book's history is attributable to them, and the more an independent examination feels like an examination of their record. The people with the most context are frequently the least willing to share it.

The individual's incentive is to protect a record. The organisation's is to see the book clearly. Those conflict directly, and no individual can be asked to resolve it alone.

The same dynamic runs at senior level, where it costs more. Asking the account owner to justify a defended balance is, in practice, asking it to qualify the growth story — and that is the number the board watches most closely. So the account that reaches the board is often the one the account owner supplied: the customer will pay, the relationship matters, the balance is old but secure. Finance is then asked to stand behind a position it was never given the data to test.

And the tenures do not match the damage. A trade channel receivable takes about five years to become unrecoverable — longer than most people stay in any one seat, on either side of the transaction. The balance is created under one team, carried by another and provided against by a third. No individual ever sees the full cycle, which is why the pattern is invisible from inside any single tenure.

Receivables decay on a five-year cycle. Careers turn on a three-year one. That gap is not a failure of character — it is why the shareholder ends up carrying what nobody else had to.

What the message has to say

Three things, stated by leadership rather than by the receivables partner. That this examines the process, not a person. That the book's condition predates almost everyone responsible for it — on one engagement 41.4% of the balance was already beyond two years the day it was handed over. And that data sharing is not discretionary — where records arrive in full and on time the work takes weeks; where they do not, it takes months and the company pays the difference.

07

Enablers and blockers

The same information request, put to two people in the same company, produces two entirely different responses. The difference is not competence. It is whether the person believes an accurate picture of the book helps them or exposes them.

Two responses to the same request
The enablerThe blocker
When asked for dataSends the full ledger, including the parts that look bad, and flags what they already know is wrongSends a summary, promises the detail, and attaches a commentary explaining what the numbers really mean
When a finding is uncomfortableAsks what caused it and what fixes itAsks whether the data is right
On a customer they defendedAccepts the ageing and revises the viewRepeats the assurance and asks for more time
On their own processVolunteers the weak points, because fixing them is the jobDefends the process, because the process is their record
What they optimise forThe company seeing clearlyA position that cannot be independently reconstructed
An enabler treats an independent look at the book as a second pair of eyes. A blocker treats it as an audit of themselves. Both readings are sincere — only one is useful to the company.

What each does to the company

Enablers compress time. Where records arrive complete and on time, analysis takes weeks and the organisation acts on findings within the first month. Where they do not, the same work takes months while the ageing continues moving in the only direction it moves when nobody acts. That difference is paid for entirely by the shareholder.

Blockers set the culture, not just the timetable. A team that watches information being withheld without consequence learns that withholding is acceptable. The behaviour spreads sideways and downward faster than any policy, and becomes far harder to reverse once it is normal.

And what each does to themselves

A person who surfaces a problem early owns the fix. A person who conceals it owns the consequence — and the consequence arrives late, at a scale nobody can manage quietly, usually in front of an audience.

Across these engagements the pattern is consistent. The people who opened their books early were the people whose standing rose. On the longest engagement in this series — seventeen years — the client's credit controller and management drove the closures, and both advanced considerably over the period.

Concealment is not a strategy. It is a loan against your own credibility, and it falls due at the worst possible moment.

How a blocker protects the position

The behaviours are consistent enough to be recognised on sight, and each one looks like diligence from a distance.

Four observable patterns, and what each does to the process
BehaviourHow it presentsEffect on the process
Knowledge held outside the systemAccount context held verbally rather than recorded, with queries routed through one person.Creates process dependency on knowledge that exists outside the formal system.
Records kept outside the systemParallel spreadsheets and side reconciliations; arrangements agreed verbally and not entered.The position cannot be reproduced from company records, so it cannot be independently checked.
Information supplied partiallyData supplied late, in summary form, or accompanied by commentary that pre-empts the reading.Findings are answered with questions about the data rather than about the finding.
Controlling what reaches managementMeeting attendance and escalation decided at the same level that holds the book.Findings do not reach the level with authority to act on them.

These are observable patterns, not stated intentions. The description below is of what the behaviour does to the process, not of what anybody meant by it. A person who has held a difficult book for years genuinely does hold context nobody else has, and genuinely did build the side spreadsheet because the system could not do the job. The behaviours begin as competence and harden into protection.

The test is simple: could somebody else run this book on Monday from what is recorded? Where the answer is no, the company does not own its own receivables process — an individual does. That is a governance exposure regardless of how well that individual performs.

Any process that cannot survive the departure of one person is not a process. It is a dependency the company has not priced.
08

Not a cost — a force multiplier

Outside assistance on receivables is usually priced as a collection cost and evaluated against a recovery rate. Both are the wrong frame, and they understate what the work produces.

Collections is the byproduct. What the process actually generates is information the company does not otherwise have, applied to decisions it is already making badly.

What the work produces, beyond the cash
OutputWhat it means in practice
A single view of the bookEvery open balance classified by cause, owner and next action — in one place, updated monthly. Most companies have a debtor ageing and no view.
Time returned to the sales organisationCommercial and field staff stop resolving paperwork and go back to selling. On maintenance and engineering books, this is frequently the largest single benefit and the least measured.
Time returned to managementOne reconciled position replaces a monthly cycle of partial answers from several functions. The review becomes a decision rather than an investigation.
A named defect list for the operating teamsNot “collections are poor” but: these forty invoices lack a proof of delivery, these twelve have a rate dispute, these nine need a credit note. Specific, owned and closable.
Process recommendations that stop the recurrenceWhere an issue category refills after being cleared, the remedy sits upstream in billing, onboarding or credit policy. Fixing that is worth more than clearing the category again.
Legacy cleared while the process is correctedThe old book is worked and the causes are fed back simultaneously, so the same book does not rebuild itself within two years.
A receivables partner that only collects has done a third of the job. The cash is this year's benefit. The process correction is every year's.

And the commercial structure is not a cost line

GGA is paid only on cash that reaches your bank. Success commission is based on cash actually received. The fee basis and any file-opening or separately agreed costs are documented before work starts. Nothing is charged where nothing is recovered. On aged balances the fee comes out of value that was, on the company's own classification, at risk of returning nothing at all.

Set that against the arithmetic above. A rupee recovered from an aged balance arrives without the sales cost, the delivery cost or the working capital of earning it again — and it removes the need to find ₹6.50 of new business per rupee lost at a 20% margin. Recovered aged debt is the highest-margin revenue in the business.

The larger effect is on cost the company has not yet incurred. Fixing the process that produced the blocked balance stops next year's book forming the same way. That is the part with no fee attached to it, and over a multi-year engagement it is usually worth more than the collections.

We are not an addition to the cost base. We are paid out of value the company had already discounted, to reduce cost it has not yet paid.

This is why the arrangement is a complement rather than a substitute. The internal credit function retains the relationship, the commercial authority and the escalation. What it gains is capacity, an independent read of its own book, and a defect list it can act on. On the longest engagement in this series — seventeen years — the client's credit controller and management drove the closures; the partnership supplied the visibility and the volume capability that made that possible.

Which places the burden of vision on the operating teams as much as on the board. A junior or middle-tier team that reads an outside partner as competition, withholds data and defends its own record is not protecting its position. It is throttling a decision the management has already taken, and doing so on a timescale where the consequences will outlast everybody involved.

09

What leadership should examine

Six questions. Each can be answered from data the company already has, and the answers are usually not known.

  1. Of last month's overdue balance, how much required an action by someone outside finance? Classify every open item by the person who must act next, and count the names that are not in the finance team. On the engagements in this paper that share ran between 27% and 78%.
  2. What proportion of our invoices cannot be validated by the customer as issued? Missing agreements, absent purchase orders, undelivered invoices, unevidenced delivery, incomplete onboarding. This is the number that determines whether collections effort can work at all.
  3. What is our invoice cancellation and revision rate? Corrections are usually handled inside billing as routine housekeeping, so the aggregate is never reported. On one engagement it reached ₹30.46 crore.
  4. Who decides when a balance moves to recovery, and is that decision written down? If the answer is a conversation with the account owner, the timing of every recovery in the business is set by the person least willing to concede it.
  5. What did our own team collect on the balances it retained, matched for age? Until this comparison exists, no claim about internal versus external performance is testable.
  6. Which balances have we known were uncollectable for more than six months and still carry? The identification and the accounting decision are separate acts. The first should not wait on the second.

The whole argument, in one line

In God we trust. All others should look at the data.

10

Method and limitations

This paper synthesises findings from fifteen GGA engagements across instruments, technology services, electronic security systems, process automation, government medical supply, agri inputs, trade channel distribution, pharmaceuticals, medical devices, project and EPC supply, logistics, leasing, publishing, testing and certification, and equipment maintenance. Each was analysed independently from the client's own line-item ledgers and monthly reconciled dashboards; individual studies are published separately with their own method sections.

The figures quoted are drawn from those studies and carry their limitations. Where a percentage describes one engagement, that is stated. No figure here is an average across engagements, because the books differ too widely in composition, vintage and sector for an average to mean anything — the consistency claimed is of pattern, not of magnitude.

Two of the engagements rest on field observation rather than invoice-level measurement and are used only where their findings are corroborated by measured engagements. The 481-account assignment test is a full-population test on one tranche of one book.

The account of incentives is inference, and is labelled as such throughout. The mechanical assignment rule is measured; the selection behaviour underneath it, and the reasons functions withhold information, are practitioner interpretation consistent with the observed pattern. They are offered as explanation rather than as evidence.

This paper deliberately reports what companies do badly. It should be read alongside the individual studies, several of which describe engagements where the internal team was the reason the result was achieved — including one running seventeen years in which the client's credit controller drove closures from inside the organisation, and one where a newly appointed finance leader's ownership was decisive.

Tax treatment is stated as a general feature of the Indian framework and varies by classification. Readers should confirm their own position; this is not tax advice.

No client, customer or individual is identified. GGA does not publish the payment behaviour of any named company.

11

Definitions

Terms as used in this paper
TermDefinition
Internal blockerAn open balance that cannot be collected until somebody inside the selling organisation acts — a document produced, an approval given, a correction made, a decision taken.
Order-to-Invoice defectA condition preventing collection that arises before the customer is involved: missing agreement, absent purchase order, undeliverable invoice, unproven delivery or usage.
Made payableThe state in which an invoice can be validated and approved by the customer. An invoice not yet made payable is not late; it does not exist in the customer's queue.
Assignment dateThe point at which a balance moves from internal follow-up to recovery. Usually the least formalised and most consequential decision in the cycle.
Routing by ownerClassifying blocked value by the function that can release it, rather than by how long it has been outstanding.
Recovery yieldValue converted to cash as a proportion of the balance in an age band. Falls steeply with age in trade channel books.
12

Frequently asked questions

Why do B2B invoices in India actually go unpaid?

Usually because the seller has not finished making them payable. Across fifteen engagements, the recurring blockers were agreements never signed, purchase orders never issued, invoices that could not be delivered for want of a billing contact, deliveries never evidenced, customer onboarding never completed, and approvals sitting undecided. On a logistics book 78.0% of blocked value required action by the seller's own functions; on a maintenance book 61.4%; on a technology services book 38.5% of the entire opening balance was blocked before any customer was contacted.

Is slow payment usually the customer's fault?

Rarely the largest part of it. Customers do refuse to pay — on one trade channel book ₹10.54 crore sat with dealers who acknowledged the balance and declined. But in every engagement examined the refusing customer was a minority of the overdue book, and treating the whole book as though it were that customer misdirects almost all of the collection effort.

Why does an ageing report fail to show the real problem?

Because it sorts by time rather than by cause. An invoice awaiting a purchase order appears identically to one awaiting payment, so both are chased by finance, which can resolve neither. Routing every open balance by the person who must act next — rather than by how old it is — changes what happens to it, and on every engagement examined that reclassification surprised the leadership team.

What happens when internal issues stay unresolved?

The balance behind them ages, and ageing destroys value. On a trade channel book worked over five years, recovery yield fell from 35.2% inside one year to 19.0% at one to two years, 9.9% at three to four, and 2.4% beyond five, with the proportion classified unrecoverable rising from 16% to 93.7%. In a channel book time removes the counterparty entirely — field visits found ₹5.81 crore owed by dealers who could no longer be located.

Why does incomplete customer KYC keep coming back?

Because the rate at which it accumulates tracks new customer acquisition, not collection effort. On a testing and certification book, KYC pending held 13.8% of collectable value, was cleared to 0.3% over nine months of sustained follow-up, and rebuilt to 9.7% over the following twelve. Every new customer arrives without an onboarding record, and where nobody upstream owns that link, a strong sales month produces a larger blocked balance sixty days later.

Who should decide when an overdue balance goes for recovery?

Not the function that created it. On one engagement, a 481-account assignment tranche was tested line by line: on 481 of 481 accounts the value released for recovery equalled the balance already beyond 240 days, with everything younger retained. Because 41.4% of that book was already beyond two years, most of what recovery received sat where yield is 8.1% rather than the 35.2% available inside the first year — so the outcome was determined before any collection work began.

Why do companies carry receivables they know are uncollectable?

Because reversal restates a reported number, the person who booked the sale has often moved on, and absent a rule forcing the question at a date it is never formally asked. On one book, ₹3.07 crore was identified as uncollectable in the first month of an engagement and was still carried in the eleventh. The cash loss has already happened; carrying it defers recognition rather than loss, while tax has been paid on revenue never received.

Why do internal credit teams resist an external receivables partner?

Because they read it as an assessment rather than as capacity. If an outside team collects money the internal team could not, that is an implicit judgement — not one anybody intends, but the obvious inference, and it is drawn before the first data file is shared. Information supplied partially, delaying handover and renegotiating scope are rational responses to a threat framed that way, and the behaviour intensifies with tenure because the longer somebody has managed a book the more of its aged balance accumulated on their watch.

Does extending customer payment terms improve a receivables book?

No. It reclassifies overdue balances as current without collecting anything. On one engagement, contractual terms on the same customers moved from 45 days to 60 with 90 under discussion, while what those customers actually paid did not change. The ageing report improves, the overdue percentage falls, and the cash position is identical. A book where terms have been extended to match how late customers already pay is one where the measurement has been moved to meet the result.

Why do finance teams accept the account owner's version of a receivable?

Because the asymmetry is structural. The commercial side owns revenue and finance owns a report about revenue, so questioning a defended balance means qualifying the growth story the board watches most closely — and finance is rarely given the ledger detail that would settle it either way. The result is that assurances — the customer will pay, the relationship matters, the balance is old but secure — reach the board carrying finance's authority without having been tested against the ledger. The tenures compound it: a trade channel receivable takes about five years to become unrecoverable, longer than most people stay in any one seat, so no individual ever observes a full cycle from sale to write-off.

How can a company test whether its own organisation is the blocker?

Take last month's overdue report and classify every balance by the person who must act next, then count how many of those names sit outside the finance team. It takes an afternoon from data the company already holds. On the engagements examined here, that share ran between 27% and 78%, and in every case the answer was materially higher than the leadership team expected.

Evidence boundary

This publication separates measured findings, modelled illustrations and operating observations. Population evidence does not predict the outcome of an individual receivable, counterparty or company.

Public law, policy terms and procedures can change. Current primary material and the facts of the live case govern.

Download the publication

Download the complete publication

Email GGA gg@ggaindia.co.in