Bad Debt: When to Stop Chasing, and What Giving Up Actually Costs
Bad Debt: When to Stop Chasing, and What Giving Up Actually Costs
Every business with receivables eventually holds an invoice that isn't going to be paid, and the decision about what to do next is almost always made emotionally rather than arithmetically. Some owners chase for years, spending far more in time and fees than the balance could ever return. Others write off too quickly, abandoning accounts that a single well-timed contact would have recovered. Both errors have the same root cause: nobody calculated the expected value. This guide provides that calculation — what pursuit actually costs, what recovery is realistically worth at each stage, why the money you've already spent is irrelevant to the decision, and how automation moves the line at which giving up makes sense.
What you'll learn
The expected value framework
The decision has exactly one correct form, and it's simpler than the emotion around it suggests:
Continue pursuing if (probability of recovery × amount you'd actually collect) exceeds the cost of the next stage of pursuit. Otherwise stop.
Three things about that formula matter more than the formula itself. "Amount you'd actually collect" is not the invoice balance — it's what you'd realistically receive after a settlement discount and after any commission, which for an agency-placed account is a substantial haircut. "Cost of the next stage" is forward-looking only — everything already spent is irrelevant, which is the single most common error in these decisions. And "probability" should come from your own history where possible: what proportion of your accounts at this age, in this situation, have actually paid?
Run through a worked example to see how quickly the math turns. A $4,000 invoice, eighteen months old, from a customer who has stopped responding. Realistic recovery probability might be modest — call it 20%. A likely settlement is perhaps 60% of face. If you place it with an agency taking a substantial commission, your expected net is a fraction of the balance: 20% × $2,400 × (1 − commission). Against that, the cost of pursuit is agency effort you don't pay unless they collect — so placement is cheap and probably worth doing. But if the next stage is litigation, with filing fees, attorney time, and your own hours, the expected value flips negative quickly, and the correct answer is to stop or settle for whatever is available.
The framework's real value is that it makes the answer different for different accounts, which is how it should be — and it removes the two failure modes at once, since it stops both the indefinite chase and the premature surrender.
Why write-offs cost more than the invoice
A bad debt is not a revenue reduction; it's a margin reduction, and the difference is enormous. The goods or services were delivered, so the costs were already incurred — the entire loss falls on profit.
| Net margin | Sales needed to recover a $10,000 write-off |
|---|---|
| 3% | ~$333,000 |
| 7% | ~$143,000 |
| 15% | ~$67,000 |
| 25% | $40,000 |
Those numbers explain several things at once. They explain why prevention — the credit policy in our net terms guide — is worth vastly more than any collection effort, since twenty minutes of underwriting can prevent a loss that takes six figures of new sales to replace. They explain why concentration matters so much in the failure curve: a business that loses two mid-sized accounts in a year may spend that entire year working back to even. And they explain why a partial settlement usually beats a write-off by a wider margin than owners intuit — recovering 60% of that $10,000 leaves only $4,000 of margin to replace instead of $10,000.
Recovery probability by age and situation
Two variables drive collectability: how old the account is and why it's unpaid. The first is well documented — recovery rates fall sharply with age as contact data goes stale, relationships weaken, other creditors get paid first, and the debtor's situation may deteriorate further. The second matters just as much and gets ignored.
| Situation | Realistic outlook | Right response |
|---|---|---|
| Administrative — invoice lost, wrong contact, approval stuck | High recovery, low effort | A reminder to the right person with a payment link. Never write these off; find them. |
| Cash-constrained — wants to pay, can't right now | Good recovery over time | Payment plan. Pressure produces avoidance; structure produces payment. |
| Disputed — believes something was wrong | Depends entirely on resolution | Resolve the dispute, don't collect it. Pursuing a legitimate dispute destroys a customer and rarely recovers anything. |
| Distressed — genuine financial trouble | Low, and falling | Settle early for what's available, before other creditors take the remaining capacity. |
| Defunct — business closed, no assets | Near zero absent a guarantee | Write off, unless you hold a personal guarantee worth pursuing. |
| Won't pay — has capacity, has decided not to | Depends on leverage | Escalation has real value here — this is the category where legal pressure works, if the balance justifies it. |
The diagnostic step most businesses skip: find out which category an account is in before deciding anything. A distressed customer and a won't-pay customer look identical from the outside and require opposite responses — one needs a fast settlement before the money runs out, the other needs credible escalation. One phone call usually distinguishes them.
Pricing the pursuit
Businesses systematically underestimate what collection costs because most of it isn't invoiced.
- Staff time. Every call, letter, and follow-up consumes hours that produce no revenue. At a loaded hourly cost, a few hours a month on one account is a real number — and one that frequently exceeds the balance on smaller invoices.
- Agency commission. Third-party collectors take a substantial share of what they recover, per the economics in our collections economy report. Reasonable for older accounts you weren't going to recover, expensive for accounts a reminder would have fixed.
- Legal costs. Filing fees, attorney time, service costs, and — the part nobody budgets — the effort of enforcing a judgment afterward, which is its own process, as our garnishment analysis details. A judgment is not money; it's permission to try to get money.
- Opportunity cost. The most expensive and least visible item. Attention spent on a two-year-old account is attention not spent on newer receivables where recovery odds are far higher — and since collectability decays with age, that misallocation quietly manufactures the next set of write-offs.
- Relationship cost. Aggressive pursuit of a customer who might have bought again, or who talks to others in your industry, has a price that doesn't appear anywhere in the ledger.
The sunk cost trap
The most common reason businesses over-pursue is the sentence "we've already put too much into this to walk away now." That sentence is exactly backwards, and it's worth stating the correction plainly: money and time already spent are gone regardless of what you do next. They cannot be recovered by spending more. The only question that matters is whether the next dollar of pursuit returns more than a dollar.
Two related distortions worth naming. Anchoring on face value — treating the invoice amount as the thing at stake, when the actual stake is expected recovery, which may be a small fraction of it. A business refusing a $6,000 settlement on a $10,000 invoice is comparing $6,000 to $10,000; the honest comparison is $6,000 to the probability-weighted value of continuing, which may well be under $2,000. And principle — the sense that letting someone get away with it is unacceptable. That feeling is legitimate and expensive, and it deserves an honest accounting: if you're choosing to pursue an account with negative expected value because it matters to you, that's a defensible choice, but it should be made knowingly rather than disguised as a financial decision.
Settle, escalate, or write off
Three paths, with reasonably clear boundaries once the framework is applied.
Settle when the debtor has some capacity and the expected value of full pursuit is below the settlement offer — which is most accounts past a certain age. Practical rules: offer a plan before a discount, since many customers can pay in installments what they can't pay at once; get every term in writing before money moves, specifying the amount and that it resolves the balance; and take payment by traceable method. The mental frame that helps: compare the settlement to the realistic alternative, not to the invoice.
Escalate when the balance is large enough to justify the cost, the debtor demonstrably has assets or income, and the situation is "won't pay" rather than "can't pay." Agency placement is a low-risk escalation since commission is contingent. Legal action is a real investment and should be modeled with the enforcement stage included — a judgment against a debtor with no reachable assets is an expensive piece of paper.
Write off when expected recovery no longer covers pursuit cost. Do it cleanly: record it properly, stop spending attention on it, and — importantly — don't take it personally or let it change how you treat good customers. A written-off account can still be revived if the debtor's circumstances change, which is why keeping the record rather than deleting it is worth doing.
Accounting and tax treatment
The treatment depends on your accounting method, and the distinction matters more than most owners realize.
Accrual accounting: you recorded the invoice as income when issued, so you paid tax on money you never received. Writing off a genuinely uncollectable receivable generally allows a deduction that corrects for this. You'll need documentation showing the debt was legitimate and that you made reasonable collection efforts — another reason to keep records of every contact.
Cash accounting: you never recorded the income, so there's typically nothing to deduct. The loss is simply revenue that never arrived. This surprises owners who expect a tax benefit that isn't available to them.
Two further points. Businesses that carry meaningful receivables should maintain an allowance for doubtful accounts rather than recognizing losses only when they crystallize — it produces more honest financial statements and prevents the lumpy surprises that distort a year's results. And write-off is an accounting decision, not a legal one: writing a debt off internally doesn't extinguish the obligation, and you can continue pursuing or later collect on an account you've written off. Confirm the specifics with your accountant, since documentation standards and timing rules matter.
Building the policy so the decision is automatic
The best outcome is not making better individual decisions — it's not having to make them individually. A written policy applied consistently removes the emotion and, more importantly, ensures the early-stage accounts where recovery is genuinely likely actually get worked.
- Define stages by age with defined actions — automated reminders in the first weeks, personal contact at defined points, agency placement at a set age, legal review only above a balance threshold.
- Set a minimum pursuit threshold. Below some balance, the cost of anything beyond automated contact exceeds any realistic recovery. Knowing that number prevents spending $400 of staff time chasing $250.
- Set a write-off trigger — an age, a balance, or a situation category — so accounts don't linger indefinitely in a queue nobody reviews.
- Require the situation diagnosis before escalation, so disputes get resolved and distressed customers get settlement offers instead of both receiving demand letters.
- Automate everything before the human stages. This is the change that moves the whole curve: when every account gets consistent early contact automatically, far fewer reach the age where the write-off calculation turns negative.
- Feed failures back into credit policy. Every write-off should update how you underwrite similar customers — the twenty-minute check that would have prevented it is documented in our terms guide.
The underappreciated point: automation doesn't just reduce cost, it moves the decision boundary. When early-stage collection costs almost nothing per account, the expected-value calculation stays positive far longer, and accounts that would have been abandoned as uneconomic to chase manually remain worth working. The businesses that write off the least are usually not the ones that chase hardest — they're the ones whose early process is systematic enough that fewer accounts ever become bad debt.
Work every account, so fewer become write-offs
HL Hunt AI Debt Collection runs the full early-stage recovery sequence automatically on every invoice — segmented by why the account is unpaid, with self-service payment and installment plans in every message — so the accounts that quietly age into write-offs get worked while recovery odds are still high.
Frequently asked questions
When expected recovery no longer exceeds forward pursuit cost. Because collectability decays with age while costs don't, most receivables cross that line somewhere between six months and two years.
It comes out of margin, not revenue. At a 7% net margin, a $10,000 write-off requires roughly $143,000 in new sales to replace — which is why prevention beats collection.
Usually, if the alternative is realistic. Compare the settlement to probability-weighted recovery net of costs, not to face value. Get terms in writing before any money moves.
Generally yes under accrual accounting, since the income was already recorded; generally no under cash accounting, since it never was. Documentation of collection efforts matters — confirm specifics with your accountant.
Key takeaways
- The decision is arithmetic: continue if probability-weighted recovery exceeds the cost of the next stage of pursuit, and stop otherwise.
- Write-offs hit margin, not revenue — at a 7% margin, $10,000 lost requires $143,000 of new sales to replace.
- Diagnose why the account is unpaid before deciding; distressed and won't-pay customers look identical and need opposite responses.
- Ignore sunk costs and stop anchoring on face value — the stake is expected recovery, not the invoice amount.
- Price pursuit honestly, including opportunity cost, which quietly manufactures the next set of write-offs.
- Automating early-stage collection moves the decision boundary, so fewer accounts ever reach the point where writing off is correct.
Stop choosing which invoices get chased
Most bad debt is created by inattention rather than by refusal. HL Hunt AI Debt Collection works every account on schedule under your own brand, with compliance enforced automatically and reporting by segment and age — so the write-off decision is made on the accounts that truly warrant it.
This guide is educational and does not constitute accounting, tax, or legal advice. Bad debt deductibility, documentation requirements, and collection rules vary by circumstance and jurisdiction; consult your accountant and, where litigation is contemplated, qualified counsel.