Outsourced collections almost always gets sold with a number: "we recover X% of your portfolio." That line is convenient for closing a contract and it's everyone's first mistake. Nobody who understands collections can promise a recovery rate without knowing the portfolio, and whoever promises it is gambling with the client's result.
Outsourcing collections works when you agree to measure what the operation actually controls —contact, the work, and its quality— and stop promising what depends on the portfolio. That's the difference between an agreement you can meet and one that starts out broken.
Why a recovery rate isn't a promise
Recovery depends on three things and the operation controls only one. The first is portfolio quality: how old the delinquency is, how many times it was chased before, whether the contact data is alive or dead. The second is the debtor's real ability to pay, which no amount of chasing creates. The third is the work itself: how many times contact was attempted, with what script, toward what agreements. Only the third is in the hands of whoever collects.
Promising a recovery rate is promising on all three. It's like asking a doctor to guarantee a cure without seeing the patient. A serious provider promises work and quality of work; the result is estimated, not guaranteed.
The portfolio rules: delinquency age decides almost everything
Before talking targets, you segment the portfolio by delinquency age. A thirty-day account and a three-hundred-day account aren't collected the same way and don't recover the same way. As a rule, the older the delinquency, the lower the probability of payment and the more expensive each attempt. Blending the whole portfolio into a single number hides that.
So the first task of outsourced collections isn't to call: it's to understand what was received. Early, mature and charged-off portfolios are three different operations, with different scripts and different expectations. An agreement that doesn't distinguish delinquency bands is measuring wrong from day one.
What you can measure: contact
Contact is the base and it's what the operation really controls. Three indicators matter here:
- Coverage. What share of assigned accounts got at least one real contact attempt in the period. An unworked portfolio can't recover.
- Right-party contact. A ringing phone is not the same as the account holder answering. Effective contact with the right person is what opens a negotiation.
- Data quality. How many phones and emails are alive. A portfolio with dead data caps the ceiling of any effort, and that has to be reported, not hidden.
What you can measure: promises and kept promises
Once there's contact, the next measure is the payment arrangement. Two numbers in a chain: how many effective contacts end in a promise to pay, and how many of those promises are kept on the agreed date. The second matters more than the first. A team can inflate promises by agreeing to anything just to log the arrangement; what can't be inflated is the promise that actually gets paid.
A promise to pay that isn't kept isn't good work; it's a promise counted twice.
That's why an honest collections dashboard separates the promise made from the promise kept, and tracks the second over time. That's where you see whether the work truly converts or just fills a report.
What you can measure: quality and compliance
Collections touches personal data and contact with people, so quality isn't only how much you recover but how. Work that recovers today but leaves complaints, off-hours contact or undue pressure costs the client its reputation tomorrow. It's worth measuring respect for reasonable contact hours, the absence of abusive treatment, and a complete record of every action.
In Colombia the handling of a debtor's data falls under the habeas data and data-protection regime, and there are rules on how and when you may make collection contact. A provider that ignores that passes a legal risk to the client. This is general framing and not legal advice; each portfolio should be reviewed against its own contractual and regulatory frame.
What not to promise
The short list of what no honest provider should guarantee: a fixed recovery rate, an exact timeline to recover, or one target for the whole portfolio regardless of delinquency. All three sound good in a proposal and all three end in a tense relationship when the portfolio doesn't yield what was promised. The reasonable promise is one of effort and method: coverage, effective contact, quality of work and transparency about what the portfolio allows.
How a sensible agreement is set
A good collections agreement looks like any SLA you can meet: it measures activity and quality, not guaranteed results. It defines minimum coverage, contact windows, record quality and a report that separates the promise made from the promise kept. A recovery estimate exists, but as a projection you revise by delinquency band, not a number carved in stone. And as in any outsourced operation, it's worth agreeing up front on what you'll measure before debating how much you'll recover.
Where smartBPO fits
In the collections we run we start by reading the portfolio, not by promising a number: we segment by delinquency, check data quality and agree with the client on what's workable and what isn't. We measure coverage, effective contact and kept promises —not just promises made— and we log every action with the care that handling a debtor's data demands. We don't guarantee a recovery rate, because doing so would mean selling a figure we don't control. What we offer is steady, measurable, respectful work, and an honest report on what the portfolio actually allows.