Allocation strategy and decisioning

How treatment choices are made across internal recoveries, DCA, debt sale, and abandon

What this decision is not

Allocation strategy is not a one-time routing rule you type into a workflow engine and forget. It is a portfolio and account-level discipline that combines expected value, operational capacity, conduct risk, legal constraints, and vendor economics. The same nominal “contingent DCA” path can mean different things depending on segment, vintage, balance band, hardship density, and whether you are on first or second external placement.

It is also not the same question as “which DCA firm won the RFP.” Vendor selection answers capability and commercials. Allocation answers which debts should enter that channel at all, in what order, and when to exit.

The choice set: internal, DCA, sale, abandon

At a high level, every serious model needs explicit rules for four families of action (with variations inside each).

Internal recoveries. The CP keeps work in-house or in BPO that runs first-party or tightly controlled processes. Reasons include high touch or vulnerable customers, litigation or complex disputes, data sensitivity, strategic relationships, or simply that internal unit economics beat external fees for a segment.

Contingent DCA. External agency works for a success fee; CP remains owner of the receivable. Reasons include scale, channel coverage, geographic reach, and variable cost. This pack focuses heavily here, especially post write-off, because volume often concentrates in that segment.

Debt sale. Receivable leaves the CP’s balance sheet (subject to deal structure). Reasons include capital relief, certainty of cash, removal of operational load, or poor marginal return on further collection spend. Sale is not “DCA but faster.” It changes legal position, customer communication, conduct oversight, and reporting.

Waive, abandon, or strategic non-pursuit. Terminal paths where the CP decides not to allocate marginal effort or risk. This is governed, not informal. See End states and closure decisions for abandon versus forgiveness framing.

NPV in this context

Net present value in recoveries compares discounted expected cash (and sometimes non-cash benefits) minus expected costs across realistic time horizons. Costs include internal FTE, DCA commission, legal, trace, letter, and the opportunity cost of management attention.

In theory, you rank options by NPV and pick the highest. In practice, three forces constantly interrupt that spreadsheet story.

First, uncertainty: early-stage models rely on historical recovery curves that may not hold for a new segment or a new regulatory setting. Second, conduct and reputation: a marginally positive NPV path can be unacceptable if contact intensity or channel choice harms vulnerable customers or breaches policy. Third, capital and accounting: sale timing and loss recognition can matter as much as raw collection cash to treasury and reporting stakeholders.

Good programmes document where NPV is the primary driver, where it is a constrainted optimiser under conduct floors, and where it is overridden entirely (for example remediation-led forgiveness).

Conduct overlays and policy floors

Examples of overlays that routinely change allocation even when raw economics favour external collection:

• Hardship or vulnerability flags that route to specialist internal teams.
• Complaint or dispute holds that block outbound until resolved.
• Credit reporting or disclosure constraints that change what can be said or when contact is allowed.
• Settlement authority limits that make DCA negotiation ineffective without frequent escalations.

Overlays should be explicit in policy and visible in data feeds. Silent overrides produce “ghost strategies” where operations think an account is in DCA treatment while compliance expects suppression.

First placement versus second placement

First placement (R1) is often about coverage, speed, and establishing contact. Data is fresher, customer recognition of the debt may be higher, and operational breakage (bad phone numbers, wrong balances) shows up early.

Second placement (R2) is rarely “same again.” It follows a recall from R1 with a reason code, a possibly different balance, a cooling-off or strategy change, and a new commercial deal. Underperformance in R1 is not automatically solved by R2 unless you change something material: segmentation, channel, settlement authority, trace intensity, or vendor.

Treating R2 as a rubber stamp is a common anti-pattern. It burns customer goodwill, duplicates contact, and destroys comparability in reporting unless trace fields are clean. See R1 / R2 traceability.

How strategy changes over time

Early after write-off or charge-off, you may prioritise breadth: get eligible accounts onto a controlled external path with strong reconciliation. Later, you may shift to depth: smaller high-balance segments, legal pathways, or debt sale for long tails.

Seasoned portfolios also face fatigue effects: customers contacted many times without resolution need a different hypothesis, not more identical calls. Governance should force periodic revisits of segment definitions, not only monthly scorecards on unchanged cohorts.

Why not everything goes to DCA

Sending all eligible debt to DCAs sounds simple. It usually fails because it ignores marginal return, conduct risk, data quality, and vendor capacity. Some accounts are cheaper to resolve internally. Some should never be outsourced due to legal or privacy constraints. Some would overload DCA intake and break SLAs, creating invisible queuing that shows up months later as complaints and missed targets.

Why not everything is sold

Debt sale can be excellent for the right portfolio and buyer market. It is not universal. Buyers price risk, data quality, and brand sensitivity. Poorly documented placements or weak chain of customer communications reduce price or kill deals. Some CPs also retain reputational or regulatory reasons to keep certain debts in-house even when sale NPV looks fine on paper.

Making decisions auditable

Strong operating models store not only what was decided but why: strategy version, segment definition, exclusion rules, and the evidence pack for overrides. That supports second line challenge, audit, and post-incident review without reconstructing emails.

Next: Written-off allocation · DCA operating model · Back to pack home