Chatbots can drastically reduce the number of interactions between customers and support agents. They increase resolution rates by 15% to 20%, delivering massive time and cost savings.
While this might sound all positive, there is one challenge businesses face when applying artificial intelligence (AI) to the customer support model: knowing when a customer's needs would be better served by a bot or a human.
Before getting to that, it is imperative to understand why AI is key to delivering a great customer experience. Good service is no longer good enough amid rising customer expectations, says Arihant Jain, Head of Business for MEA at Freshworks. For Jain, AI can be a significant differentiator for businesses to engage consumers in new ways.
"Today's consumers expect omnichannel contact options. They also want conversational self-service interfaces and fast resolutions. On the one hand, customers are demanding unique, innovative and special service. On the other, businesses want to justify their investment in great customer service. AI can help in both instances," says Jain.
There's a lot to think about when delivering a coherent experience across all touch-points. Customers want to contact businesses on their own terms and on familiar channels. These include call centres, social media, e-mail, Web forms, physical stores, online chat and apps. They also want to seamlessly switch between channels and continue their conversation from where they left it.
To deliver a coherent experience, businesses must uncover patterns across an overwhelming number of data points. This is where AI makes a real difference. AI can consume vast amounts of data at far greater speeds than humans, and can learn from interactions," says Jain. "Since AI systems can see, talk and hear, customer experience (CX) teams are now entering a new era of creating AI-powered experiences that feel like natural human engagement."
Once the need for AI is established, it then comes down to businesses finding the right balance between knowing when it's appropriate for bots to serve customers and when an issue requires a human touch.
"AI meets three critical needs in the age of the customer: personalisation, contextual intelligence and immediacy. However, businesses need a hybrid CX model that combines the ingenuity and efficiency of AI with the empathy of humans," says Jain.
Bots are good at handling first-level queries. For example, they can answer simple, common questions, check order statuses and process payments. By automating responses to common queries, companies can reduce training time for service reps and employ fewer people to handle repetitive queries. This decreases call handling times and costs, and increases the number of first-call resolutions.
While first level query handling is one part of the solution, bots can also assist human agents.
For instance, bots can route inquiries, interpret incoming messages, recall important information and develop initial responses. Service agents can then edit these and decide how to weave the information into conversations with customers. "This often shortens call wait times and call handling times," says Jain.
Here are a few scenarios where a chatbot might escalate a problem to a human:
* When the request is not understandable or is better served by an agent (eg, conversion or attrition);
* If the customer appears frustrated;
* If there is no self-service option;
* For high-value transactions and sales opportunities; and
* The customer explicitly requests a human agent.
Challenges and opportunities
While AI promises to supercharge the customer experience, there are challenges in applying AI to customer support.
"High deployment costs, which impact return on investment; data challenges, such as integration, privacy and regulatory obligations; cultural change management; the speed at which technology evolves; and the resultant lack of skills or knowledge, could all hamper AI efforts," says Jain.
Executive buy-in and integrating AI into business decision-making are crucial for success.
If done right, AI has the potential to create a frictionless customer experience, delivered at scale and speed.
Opportunities that arise out of this include:
* Access to insights that can help support teams make informed decisions that impact CX, such as refining their overarching CX strategy and providing real-time responses to customer inquiries.
* Automation of behind-the-scenes processes, such as targeted marketing promotions and messaging.
* Better customer interaction with chatbots to have their queries answered immediately rather than waiting in a call centre line.
Before engaging in AI, businesses should ask three key questions. What is the purpose of AI in the enterprise? What business and customer problem are we trying to solve? And how will AI improve the customer experience?
Here's a step-by-step plan to implement AI in the enterprise:
* Assess the AI capability (skills and data) in the organisation;
* Identify the leadership to drive the AI strategy;
* Craft a vision for what the organisation wants to achieve with AI;
* Explore AI's initial, high-value use cases and the technologies needed to implement them, then build capabilities to develop those use cases;
* Launch proof-of-concept and pilot implementations on selected use cases;
* Scale the pilots to business-wide scope;
* Establish governance to prioritise AI projects; and
* Nurture an AI/insight-driven culture.
AI will play a big role in CX going forward.
"AI could improve response times, provide relevant, personalised recommendations, incorporate sentiment into responses, eliminate bottlenecks and automate routine inquiries. This will free up humans to deal with more complex problems," says Jain.
All said and done, Jain emphasises that a blended approach, one that combines man and machine, is key to avoiding mistakes.