About
Discovery innovates with a Connected Customer Services Platform for superior CX. 3.9m customers benefit from advanced data insights and AI, yielding 8.9 CSAT, 85% CES, and 10% call/email rate reduction, with bespoke CRM that drives ongoing success.
Innovation presentation
Discovery Health is a leading medical scheme administrator in South Africa, providing administration and managed care services to over 3.9 million members, with a market share of over 58% of the open medical scheme market in South Africa. We manage 18 restricted medical schemes on behalf of leading corporate clients, as well as Discovery Health Medical Scheme, thus managing 39% of total membership of South African medical schemes.
Discovery Health customers have access to a variety of servicing channels, to ensure we meet client expectations where they need us. We receive over 45 million customer interactions across servicing channels per year, and the usage is carefully monitored, alongside the customer experience within each channel.
We are obsessed with our customers and their experiences with us because we empathise as though their problems are our own and recognise the impact of our actions on the people we serve – especially in the healthcare industry where people are at their most vulnerable. Our focus is placed on providing a personalised service offering that enables moments of delight, supporting each customer based on individual circumstances. We predict potential problems before they arise, and proactively intervene to solve customer challenges before they realise it is there.
To enable this level of sophistication in service delivery, Discovery Health has placed significant focus on integrating data science capabilities within service operations. The introduction and integration of these capabilities were achieved through the following core deliverables:
1. Building Data Models, Indices and Repositories
The Data Science Lab has 8 Data Scientists who support operations and servicing, dedicating their time to building purposeful models that make a significant difference in the lives of our customers. Deep analysis of customer journeys at scale requires well curated data sets. Our own journey to doing such detailed analytics began with the design of our most comprehensive data asset – the Interaction Data Repository (IDR).
The IDR not only stores a single timeline of all customer interactions, but also augments these with any additional data we have available – such as service agent details, quality scores, and bespoke natural language processing model results in the form of sentiment, emotions, and topics.
The IDR forms the basis for a collection of indices; single personalised metrics per entity which characterise different elements of their behaviour. Examples of these include a digital index, contactability index, affinity index, most frequent channel index, service load index, service storm index, sentiment indices etc. The indices are used within our servicing CRM system across all channels and for data science modelling across the business. The indices are updated as often as daily and the creation of these data assets has driven the quality and collection of all operations data. To this end, the IDR forms the basis of the next-best-action recommender models acting as a consistent single source for all reporting in future.
Today, 32 bespoke-built customer insight data models are already operating across Discovery Health service channels, adding sophistication and smarts to service rendering – often without the knowledge of customers, importantly our customers and agents reap the benefits of superior service delivery, evidenced through performance results.
Future servicing excellence will depend on our ability to understand cross-channel integration, personalising the content of interactions using generative AI and providing business with predictive insights.
2. Service Operating Platform with bespoke CRM
We have built a new servicing operating platform that brings together modern customer servicing software (CRM) with our core systems as well as the power of data and the service operations environment. We call this our Connected Customer Services Platform. Architecting and equipping the platform that supports our contact centre, has been key to delivering an insight led operation, enabling data-as-a-service to provide the competitive advantage we seek.
We are now in a position where data-as-a-service is well on course, with expansive analytics, ease of model creation and deployment into the real world. Many models are operating and aiding the business to service smarter, faster and helping team members approach each interaction with rich context about the client, ranging from simple NLP-based sentiment tracking and service intervention nudges in near real-time, through to far more complex models that identify and predict clinical needs that trigger proactive service activities in a personal and meaningful way, impacting the quality of health and life of our customers.
Our new CRM (E-Volve) is a bespoke build for Discovery Health. This system and its architectural design, enables easy provision of the intelligent personalised insights to all our team members interacting with our customers. It is designed to integrate easily with data assets, creating a powerful platform for our data scientists to bring data driven intelligence to life in day-to-day activities with our customers, adding behavioural as well as clinical intelligence to servicing.
Within our CRM, we have built an interpretability layer. This alerts agents to the most pertinent data science indices available for the customer and suggests how best to approach the interaction. As more models are incorporated, service agents can focus on what is most relevant and important to the customer.
We are transforming contact centre design by investing differently in the expertise, people and technology, simultaneously we’ve taken the two-dimensional traditional CRM service platform (most customer service leads are saddled with) to the next level. This new three-dimensional capability provides far more colour and insight about each of our customers during service interactions, enabling personalisation and meaningful touch-points that help us leave a memorable impression with customers and deliver on our purpose – to make people healthier, and enhance and protect their lives.
3. Enabling expansion AI Capabilities for service and operations
With the significant advancements in artificial intelligence capabilities, Discovery Health has partnered with leading industry experts, and developed, tested and deployed 3 custom-built AI components to support service operations, integrated into the Connected Customer Service Platform:
3.1. Virtual Expert
Merlynn’s Tacit Object Modeler (TOM) replicates human expertise. TOM learns directly from the human expert to understand their decision-making process and requires no historical data to learn. TOM can replicate complex decisions including tacit knowledge - knowledge gained through experience which is very difficult to explain or articulate and is often described as “gut instinct” and intuition. The technology is used to digitally scale human expertise enabling organizations to better manage risk. Digitized expertise provides consistent, unbiased, accurate decisions, reflecting the expertise of the organization’s top decision makers. Decisions are traceable and transparent. We have used this Virtual Expert to replicate the decisions made by deep subject matter experts in our business, most notably our Discovery physicians and medical advisors with use cases including complex clinical cases which fall outside protocols and standard guidelines.
3.2. Discover It (Gen A.I.)
A custom-built platform, leveraging the capabilities of PaLM2, LlaMA, Gemini and ChatGPT, creating an internal chat-bot for agents to easily reference product and process information in real-time, and becoming an extension of our knowledge management capabilities. We have ingested our legacy knowledge repositories and other key information sources and delivered the Gen. A.I capability to our contact centre servicing staff. It means less time finding answers, coupled with greater precision (not to mention to the point information) instantly. This service is accessed as a stand-alone and through the Evolve CRM – acting as part of the Connected Customer Service Platform. This brings improved speed and accuracy to query handling across channels. In parallel a version for our customers is underway, with integration into our core systems for absolute personalised servicing.
3.3. Custom GPT for Customer Communication
Rewriting all customer communications with the help of custom GPT, supporting our brand feel, improving content style for readers, and simplifying customer communication. This GPT has been trained on the Discovery Health brand tone, brand style, the companies values, purpose and approach, so that it can support document writing going forward as well as rewrite all existing documents/correspondence to be more simplified, engaging and better meet our intended customer centric tone
Our focus is geared towards our ambition for our contact centre to be the very best in the world, future fit, with key differentiators in customer loyalty, satisfaction and engagement. Our ability to integrate data science capabilities within operations continues to enhance our customer and employee experience.
(more information available in the supporting documentation)
Uniqueness of the project
The uniqueness of our project lies in the transformative ecosystem, where data science converges with operations’ software, efficiency, people and customer-centricity. This powerful link redefines the boundaries of possibility.
1. Agile Model Deployment
The ecosystem seamlessly deploys new models in the operational environment. This agility enhances efficiency and enables rapid adaptation to evolving customer needs. New models are easily identified, developed, tested, and deployed, enabling us to adapt to evolving customer needs, and do so swiftly.
2. Data-Driven Customer Experience
By harnessing data, we’ve personalized and anticipated customer interactions. From pre-emptively addressing potential issues to tailored recommendations, the ecosystem makes it possible to exceed expectations.
3. Holistic Integration
Beyond sophisticated data science, the holistic approach integrates these capabilities into every operational facet. This reshapes both customer service and our employee experience and sets a new standard in health insurance excellence.
4. Collaboration
The synergy between operations and data science teams fuels analytics success. Agents receive data science exposure early on, while data scientists explore novel ideas.
This unique ecosystem transcends the sophistication of algorithms; it amplifies the impact of our data science capabilities and changes the job to be done by contact centres. We have forged a powerful link where data science meets operational efficiency and customer-centricity, resulting in a truly unique endeavor that redefines the boundaries of possibility.
