Can Predictive Analytics Software Really Determine Who is Going to Pay?

Predictive Analytics is Helping Collection Agencies Predict Who is Most Likely to Pay

For decades, collections centers have relied on the tried and true method of cold-calling as many debtors as possible and relying on the skills of individual phone agents to drive payments.

However, every year, staying successful using the same old methods gets harder and harder as legislative measures such as the Fair Debt Collection Practices Act are passed that put more restrictions on how collections call centers can operate.

To continue to grow success as a collections agency, you need to be able to liquidate as many accounts as you can as quickly as possible. This requires more efficient and targeted calling practices that focus on debtors who are going to pay once you’ve identified your Right Party Contact (RPC).

But, how can you know who is and isn’t going to pay? One highly reliable method is to use voice analytics software to analyze your first contacts with debtors to determine who is going to pay.

It sounds fantastic, like something out of a science-fiction story. But, predictive voice analytics software can assess how likely a given debtor is to pay after initial contact, and rank all of your first contacts by likelihood of payment on a follow-up call.

Here’s how it works:

Recording Data

Predictive voice analytics begins with a basic staple of the call center industry: a phone recording. Virtually every call center in the collections industry records their call for a variety of quality assurance and compliance purposes.

Predictive voice analytics software takes the raw audio of a debtor’s speech and breaks it up into specific voice feature data using digital signal processing. Examples of voice features that are quantified by this process, include vocal pitch, resonance amplitude, tone and frequency, in addition to the emotions and behaviors expressed by the debtor.

Turning Voice Data into Predictive Analyses

Once the voice data has been recorded and processed, it’s plugged into an advanced machine learning algorithm to create an objective, mathematical assessment of the speaker’s emotional behavior and tone.

This math-based assessment of a speaker’s tone is much more reliable than the “gut” feeling of most phone agents, as it is based on repeated observations of call outcomes. You see, this analysis isn’t carried out once in a vacuum; it’s performed over and over again for thousands upon thousands of calls.

Each new call recording gives the algorithm more data and real-world outcomes to compare and contrast. With each comparison of emotional state to future call result that the algorithm acquires, the accuracy of the predictive model increases.

You probably have some call center employees with:

  • decades of first-hand experience in assessing how people react to collections calls;
  • terrific recall to remember the calls they’ve worked; and
  • excellent logical reasoning skills largely untainted by personal bias.

But even your best agents don’t have perfect recall or flawless logic – and everyone has good days and bad days.

What about your newly hired agents? Or those agents that are currently struggling?

With predictive voice analytics software, you get an artificial intelligence that simulates all of the traits of your best agents on their best days. Then it applies that intelligence across 100% of your calls for each of your agents.

By assessing thousands of calls in mere minutes, the machine-learning algorithm can accrue more experience in assessing your target customers in a few hours than most phone agents will in their entire careers.

Using accrued experience of tens of thousands of calls, a voice analytics program is able to rank your contacts with debtors, creating a predictive analysis of how likely they are to pay on a follow-up call.

Proven Results

Computerized analysis of collections calls isn’t some bit of science fiction, it’s a scientifically proven reality that is already being used by industry-leading collections agencies.

In fact, users of RankMiner’s predictive voice analytics solution have increased their call center performance by as much as 63% by focusing efforts on targets identified by the RankMiner software.

If you still need convincing, check out the results that predictive voice analytics have provided for other collections agencies.

 

Download RankMiner's White Paper Measuring and Improving Customer Profitability