For instance , financial services or healthcare companies may use AI to track and detect moments of customer embarrassment or surprise when it comes to paying bills or overdraft fees . They can then provide the appropriate coaching to agents to use the best wording to evoke less embarrassment or surprise to optimize for the customer ’ s comfort .
5 . PERSONALIZING THE CX By using AI to collect , aggregate , and analyze customer interactions across touchpoints — from digital browsing activity to customer feedback and transactional data — brands can establish a clear picture of customer behaviors and interests .
They can then use these insights to personalize content and experiences for customers across channels to serve up information that ’ s relevant to the customer and their individual journey .
6 . STREAMLINING CUSTOMER
SERVICE RECOVERY Using AI to extract customer insights from across interactions and touchpoints can help brands understand why customers are reaching out to customer support in the first place , so that teams can provide better service .
These insights can be used to improve customer support queue management and can be shared with agents to keep them in the loop about who their customers are and why they ’ re reaching out . This helps agents to communicate with customers with greater knowledge and empathy .
AI can also be leveraged to provide agents with automatically generated recommended next-best-actions to take based on a customer ’ s history with the company .
7 . PROACTIVELY CLOSING THE LOOP
WITH CUSTOMERS AI modeling can be used to predict signs of customer churn . For instance , if a customer calls repeatedly without getting their issue addressed or exhibits high levels of frustration during a contact center interaction , companies can set up alerts to notify supervisors or a dedicated team to intervene .
A member of the dedicated team or a supervisor can then proactively reach out to the customer and provide a resolution for the issue , offer a special promotion , or give the customer the chance to voice their frustrations . Interventions that can drive retention and loyalty .
THE ROLE OF HUMAN ANALYSIS ALONGSIDE AI AND ML
While AI and ML can help companies save time , these technologies won ’ t eliminate the need for human involvement . In fact , human intelligence and experiences are crucial to the successful adoption of these tools .
That ’ s why savvy organizations take a hybrid approach , relying on professionals to train these models to work more accurately and responsively , as well as to monitor and respond to things that truly matter to the business .
Humans bring important business context to the table that AI and ML tools simply don ’ t offer insight into , such as knowledge of changing corporate goals or learnings from strategic planning sessions .
For instance , retail staff might ensure these technologies are set up to flag critical conversations about Black Friday leading up to , during , and immediately following the event .
As another example , models powered by AI can be set up to track and create alerts when customers use language that signals an emergency , such as a natural disaster , so that the right folks can be looped in immediately to take swift action .
While a tool might be used to route conversations that include words like “ dying ” or “ fire ” to the right team members , human oversight is necessary too .
That ’ s because people can use these very same words in a completely different context to describe something that ’ s positive , funny , or great . People
ARTIFICIAL INTELLIGENCE
might also use these words hyperbolically to describe a poor CX .
If that ’ s the case , contact center professionals reviewing these alerts can make sure they are de-escalated ( if they ’ re not being used in response to a true crisis ) and get routed to the right individuals , such as to a customer service recovery team member instead of a disaster response team member .
AI is getting smarter , and as it becomes more sophisticated , companies will be able to automate more tasks , meaning less human intervention will be needed .
As it stands today , however , human interpretation is almost always needed to tweak what AI is monitoring , flagging , and reacting to , to enable these models to take action on the right things .
COMPANIES NEED TO KNOW WHAT OUTCOMES THEY HOPE TO ACHIEVE ...
FINAL THOUGHTS
Before diving headfirst into the world of AI , it ’ s important that contact center leaders have firm business objectives in mind . Companies need to know what outcomes they hope to achieve and what pressing questions they want answers to , rather than simply implementing a product just because everyone else is doing it .
It ’ s also important to keep in mind that change management is key to successfully adopting any new technology within the contact center . Managers must bring agents on board with not only using a new solution but adapting to changing team processes , including reporting , metrics , and performance evaluations as a result of rolling out the new tool .
Joanna Moser is the global Solution Principal for Medallia ’ s Analytics Products , including Speech & Text Analytics , Social , & Action Intelligence . Prior to Medallia , Joanna spent four years at Clarabridge as a consultant leading implementations for complex , on-premise accounts and as a lead on the reporting team .
NOVEMBER 2023 39