Traditionally, the amount recovered from debt collections is the ultimate measurement of performance metric. While this is still a key goal, lenders and collectors are now placing an increased focus on call quality as another key metric. It is no longer just about how much money is recovered, but also about how you managed to get the money from the borrower.
Companies are concerned with brand reputation, which translates into compliance and call quality in terms of collections. Quality assurance measures a call operator’s effectiveness and evaluate his/her adherence to internal policies and procedures during customer interactions. This has always been a cornerstone in Flow- aligned with our mission to promote ethical digital debt collection through customer-centric operational processes and AI technologies. In Flow, high-quality negotiation techniques such as voice-to-text quality check, emotion detection and skill-based call routing are prescribed. Operators’ productivity is evaluated to match them to specific accounts. Our collection strategies are constantly iterated and tailored to optimise recovery efficiency, increase delinquency prevention and to deliver positive experience for the borrowers.
The Data Science and Operations teams have recently completed an MVP of a quality assurance platform called EQATE, short form for Ethical Quality Assurance Tech Engine. It has been implemented in some of our operating countries. Flow conducted an interview with the two leads for this project, Przemek Januszaniec, Chief Performance Officer and Dima Sakovich, Data Science Team Lead.
Flow: What was your thought process when developing EQATE?
Dima: This was our internal initiative, how we can help our QA team be more effective and reduce the amount of routine work.
Flow: How does EQATE work and how it can help in quality assurance for Flow?
Przemek: EQATE is the tool supporting management of the call centre. It works in two layers. Internal data engine is converting recorded calls into digital DataMart using artificial intelligence and machine learning algorithms. Afterwards, Graphical User Interface is presenting the information about behaviours of the operators in an interactive way to the supervisors and quality assurance officers.
The main advantage of the tool is objective and factual assessment of the calls from the perspective of the procedures, scripts and ethical standards followed during the call.
Flow: How is EQATE developed?
Dima: We use django/vue.js as a dev stack. Also, for the audio processing functionality, we developed several mathematical approaches, everything works on python. We have our own speech recognition and sentiment analysis models, but in some cases, we also use cloud solutions.
Flow: What are you most excited about using EQATE for the operations team?
Przemek: I’d say the possibility to apply EQATE cross functionally in all business areas of Flow. The initial goal of EQATE was the assessment of following the ethical promise by call centre operators. Currently we are working on utilization of the EQATE in quality assurance, training, coaching, motivation systems, customer experience. The deeper we go into it the bigger power of data is recognized and new ideas of utilization are invented.
Flow: How did you roll out EQATE to the operations teams?
Przemek: Rollout of the system required from the team two-dimensional approach. The most difficult part was having proper coordination of the user requirements and the creativity of the developers. On the other side, implementation of the EQATE was an element of digitalization and automation of the quality assurance processes.
All of that was requiring high level of teamwork in redesigning of old business processes and introducing new ways of thinking about behavioural analysis of call centre dynamics.
Flow: How do you plan to integrate EQATE to the other collection strategies?
Przemek: Implementation of EQATE opened some new opportunities for the operations team.
It was recognized that behavioural studies may be utilized not only in reporting but also in optimizing business processes. The most natural utilization is segmentation of the operators based on objectively assessed skills and competences. Further steps may be customers stratification based on recognized behavioural characteristics recognized in the digital analysis of the recorded calls.
Quality Assurance as a Key Indicator
While lenders and collectors view the importance of improving borrowers’ experiences during collection, does that mean the other key collection metrics such as productivity and recovery rates are no longer important when managing performance? Not at all. All of these metrics remain and will always be important in measuring and driving performance. However, a balanced collection strategy and operation acknowledge the important role quality assurance plays in managing performance.
Contact centres are the first face of the company and the brand advocates for the company. It is key to note that the contact centres and people should be adequately trained to ensure world class customer service. Therefore, it only makes sense to not only motivate the employees but also optimise the company’s call centre quality assurance strategy along with enhancing the quality key performance indicator.
In summary, measuring quality is critical to performance management in a call centre as it reduces reputational risk while driving behaviours and attitudes that ultimately lead to a better borrower’s experience and a lasting, repeatable collections strategy. With these goals in mind, quality is as important, if not more than, the amount collected metric in a collection company’s key measurements.