.png)
We are thrilled to share that Ophelos won the 'Machine Learning in Credit & Collections' category at the Credit Connect Awards!
It's an amazing milestone in our quest to simplify, digitise and automate the debt collection process. And we thought we'd take this opportunity to share one of the ways we're using machine learning to help individuals and businesses take control of their financial health: by creating a customer-first communication strategy.
One-size-fits-all fits no one
One of the hardest challenges to solve within customer service is deciding who to contact, at what time, and by what method. There are many different factors to consider when creating contact strategies in collections, such as information about the customer, a summary of interactions with the customer, and conversational history.
At scale, this is a near-impossible computational problem to solve and as a result, organisations default to rules-based contact strategies, in effect treating most customers exactly the same. This has proven to be a headache for both businesses and their customers, particularly within collections.
Currently, customers are often contacted at inconvenient times in the day, through outdated communications channels such as phone calls (or even home visits), and with generic messages. This leads to customer frustration and negative service metrics such as low NPS or CSAT.
Businesses face high costs as traditional phone channels are still the most expensive (and most used) communication channels within collections. To increase efficiency, organisations often add strict wrap-up times or average call times, which prevents agents from developing rapport with customers. Additionally, monitored call times foster a negative company culture and lead to higher workforce turnover, bringing inevitable hiring costs to the business.
Bespoke communications that get results
At Ophelos, our R&D team have developed machine-learning models using our proprietary “Approximate Dynamic Programming” process which create bespoke communication strategies for each individual customer and automate the communication process through our platform.
The model outputs are ingested into our collections platform, allowing us to fully automate the outbound communication process. Every customer that we engage with will have their own individual engagement strategy - every contact point is unique, leading to a significantly improved customer experience.
The results of our approach are transformative. We have seen a 25% increase in recovery rates, a 28% decrease in outbound communication costs and a 68% increase in customer satisfaction. At scale, this approach can save billions for our clients as well as significantly improve customer relations.
To learn more about our customer-specific communications and other ways we're using machine learning, why not get a demo?
