Predictive dialer for collections: when it pays off and when it does not
A predictive dialer is not better than a progressive one: it is better above a certain size. Below that threshold it adds abandoned calls without gaining contacts.
En corto
- From about 20 agents a predictive dialer starts to pay off over a progressive one.
- With fewer agents, go progressive: zero abandonment and a steadier pace.
- Law 21.320 applies to collections, not sales, and is met before dialing.
- Abandonment is measured against answered calls, not against originated calls.
What makes a predictive dialer different
A progressive dialer places one call per available agent. Simple and with no abandonment: if no one answers, no one is left waiting.
A predictive dialer places several lines per agent, anticipating when they will free up. When the math is right, the agent finishes a call and already has another connected. When it fails, someone answers and there is no one to take it: that is an abandoned call.
Why size matters
Prediction is statistical, and statistics need volume. With 8 agents, two finishing at the same time is noise; with 40, it is a pattern.
In practice the cutoff is near 20 agents. Below that, a predictive dialer adds abandonment without improving contacts per hour, and a progressive one does the same job without that cost.
This is not a software limitation: it is how prediction works. Any predictive dialer has the same threshold.
Law 21.320, before dialing
For collections, Chilean law limits how often a person can be contacted and in which hours. The fragile way to comply is to configure lists and trust that no one slips up.
The way that works is to evaluate the rule before originating the call: if that contact already hit their quota, it is not dialed. Nothing to remember and nothing to audit afterward, because the call never went out.
An often-missed detail: WhatsApp and SMS share the same legal quota as calls. If the system counts them separately, you break the rule without noticing.
How to measure abandonment right
Abandonment is computed over answered calls, not originated ones. Measuring it against originated calls gives a much prettier and completely useless number.
The 3% cited as an industry standard is just that: a standard, not a law. It is worth having as a target, but it helps to know where it comes from before treating it as an obligation.
And the number worth comparing is not the one the CRM reports, but the one from the phone platform. We have seen an operation report 90% contact rate when the real one was 3.4%.
If you have a collections portfolio up and running, the question is not which dialer is better in the abstract, but which one pays off with your headcount and your portfolio. That is answered by trying it: 2 weeks with your real data, no card.
Probar NeervoX 2 semanas gratis