Embedding Analytics Into Customer Interactions
Last month we provided an overview to Operational Analytics and now we’d like to dive deeper into the various use cases that we talked about. Customer Service centers are going through several evolutions all at once. Similar to the evolution on the marketing side, customer service groups are being overrun with new customer channels that are typically customer-initiated and always on (chat, web, mobile, knowledge bases). As omni-channel contact centers struggle to install and adopt all of these new channels, they are also being asked to become more sales oriented to help retain and cross-sell. Meanwhile, the idea of proactive service and predicting customer needs is consuming yet more of the bandwidth from customer service organizations.
Customer service organizations are finding that the automation and processing of information is the only way they can stay afloat. Flooding customer service with information is not helpful. Agents cannot synthesize the information fast enough to formulate a plan during a phone call or a chat conversation. So customer service doesn’t just need data integration and the ability to absorb information from sales and marketing data silos. They need intelligence and advice. Customer service organizations are asking themselves the following questions:
> How can I ensure that the right agent is talking to the right customer?
> How do I differentiate service to customers by value to the organization?
> How do I know what offer a customer will really react to?
> How does my help desk know if they should offer an alternative product?
To address these evolving demands, customer service applications must have access to data that allows them to tailor a new customer experience with a different set of outcomes. Models and metrics need to be readily available to the application and displayed to customers and users in a way that makes it obvious what the right next step should be…and provide insight into why that is the right next step. Customer service agents and managers are eager to turn call centers into retention and selling centers, but are demanding the right tools to do their job. The following are core components and scenarios that are required to transform the customer service function.
Models and Scores. Large organizations tend to have analytical groups with senior data scientists and statisticians working on a variety of analytical projects. Creating these custom models results in custom scoring algorithms that are finely tuned to the organization but may take a while to build – and armies of statisticians can get expensive. New analytical applications are providing the ability to create predictions (churn, offer response, etc) and propensity models without a need for deep knowledge of statistics. Further, these applications provide English language descriptions of the data attributes that influence models as well as expose those models to third party applications right out of the box. Companies like Emcien are providing models that are easily updated and leveraged by third party applications by design. The following scenarios show how these models might be used.
Current Customers. When current customers reach out to your customer service departments, there is an opportunity to improve the relationship and the customer value. No matter which channel the customer reaches out through (chat, phone, etc) – several pieces of information can be used by that app or by the agent to maximize that interaction. Key data elements include, but are not limited to, probability in customer change (positive/negative), churn potential, next best product, next best offer, or propensity for a specific offer or product. With this information, scripting can personalize the experience for the customer. For instance, an agent may be prompted to offer the customer a specific incentive if they are at risk to churn or maybe the customer is prompted for more information (demos, interests, preferences) to help place them in a marketing program.
New Customers. New customers are typically a greenfield opportunity for organizations. New customers go into marketing nurture cycles to start driving incremental value. However, in the new scenario, we can predict the customer’s value and customer’s interests from the beginning. No longer is this a greenfield approach, but now in the customer’s first customer service interaction, we have instantaneous insight into the customer by calling out to a predictive model in real-time. For instance, we may predict that this type of customer is likely to churn within 6 months. Or we might predict that they are likely to respond to an upgraded product. The customer service agent can then respond to the customer with the right offers.
These concepts can be difficult to visualize, so please follow this link for a banking demonstration of how real-time analytics can influence a current customer and new customer interaction.
AmberLeaf (www.amberleaf.net) combines strong business and operational planning with innovative technology solutions to ensure our client base serves the right customers in the right ways to generate the greatest return. To learn more about how we can help your company improve customer experience, contact us at 312.474.6120, or email@example.com.