The March Feature, “Speaking the Customers’ Language(s)” touched on AI-driven automation to help engage with U.S. and Canadian non-English/non-French primary speakers through the contact center. In this article, we dived a little deeper into the potential of these new and developing tools to enable excellent customer experiences (CXs) with our expert panel. They are:
- Salvador Ordorica, founder and CEO, The Spanish Group
- Fabio Sattolo, Chief People & Technology Officer, Covisian
- Heather Shoemaker, founder and CEO, Language I/O
Q. What are the pros and cons of AI, and its best (or not advisable) uses when engaging with these customers? Is it ready for prime time but if not, can it ever be practical for quality customer service?
Salvador Ordorica:
AI tools are growing in adoption, but they face limitations in cultural sensitivity and contextual accuracy:
- Best Use. AI works well for answering FAQs such as order tracking or product information, or for managing basic inquiries like password resets or appointment scheduling. And even supplementing live agents by identifying questions and routing them to the appropriate department.
- Challenges. Advanced conversations requiring empathy or nuance are better handled by humans. AI has difficulty grasping idiomatic expressions or regional dialects, which can most of the time lead to miscommunication.
Fabio Sattolo:
The positive side of AI is that it can absolutely be used to increase agent productivity and reduce call handle time. In addition, it can help use customers’ time more effectively.
The downside is that while AI is ready for prime time, most contact centers are not yet using it responsibly.
Importantly, contact centers should not be using AI chatbots as a primary way to engage with customers. Instead, customer-facing AI tools should be overseen and managed by live agents. This human-first, personalized approach is the best way to involve AI in customer service.
AI is increasingly being used for translation of text and speech in real time. But we first need to determine customers’ willingness to interact in this way and not force them into conversations that they may perceive are driven by bots or computers.
This brings us to Responsible AI, which boils down to this: using AI in a way that benefits and enhances the customer relationship rather than alienating the customers.
Responsible AI should always start with a customer interaction, which can be language-specific.
After the initial positive human interaction, the agent can determine if they should transfer the customer to an AI tool in the customer’s preferred language for information gathering or administrative functions. And the customer must always have an easy option to return to the agent to preserve the relationship and build loyalty.
Heather Shoemaker:
AI translations accurately and cost-effectively allow customers to feel connected and heard with the brand. It is scalable, supporting hundreds of languages to serve a diverse customer base.
AI translations are also efficient and brand-consistent by using the same quality and type of service, no matter the channel customers use. Lastly, they comply with regulations and are secure, meeting high data protection standards.
But at the same time, AI-driven large language models (LLMs) are built by real people with unconscious biases. Which means products and services that use them (like translations) can be problematic when addressing customers from different backgrounds and who speak different languages.
“As automation grows, businesses can offer multilingual support more efficiently...” —Heather Shoemaker
AI is practical and effective. But it’s necessary to deploy these tools with human oversight and if necessary, intervention.
Q. Are non-English/French language primary/preferred customers switching to/also using more automated AI-driven interactions (IVR/ITR/chatbots) and live chat from live voice? And if so, what is the impact on the ability to engage with these customers in the language of their choice?
Salvador Ordorica:
We’re seeing more customers use AI tools like chatbots and IVR as these systems become more accessible, but their effectiveness depends on multilingual capabilities. Customers may feel alienated if their preferred language isn’t supported.
“AI tools...face limitations in cultural sensitivity and contextual accuracy...” —Salvador Ordorica
For example, we’ve worked with organizations to integrate AI-assisted tools in Spanish and Tagalog, which led to a 20% drop in call volumes while maintaining high customer satisfaction. This success came in hand with technology and our customers’ culture and language customization.
Fabio Sattolo:
We do see more non-English language customers using automated AI-driven interactions. But again, the key is allowing a human agent to engage first and then determine the best time to leverage the AI tools.
Engaging with customers in the language of their choice conveys respect and a willingness to go that extra mile when a customer is already experiencing frustration or some sort of service challenge.
“Responsible AI should always start with a customer interaction, which can be language-specific.” —Fabio Sattolo
Customer support must reposition itself as a valued service, rather than try and limit human interaction or engage with customers only in one language.
Heather Shoemaker:
AI-driven interactions are shifting more towards automated customer service solutions like chatbots, IVR, and live chat, making it easier for non-English-speaking customers to access their favorite brands and consistently engage in their preferred language across channels.
This shift impacts how organizations address accessibility and customer preferences and personalization, as well as reducing reliance on live agents who may struggle with translations. As automation grows, businesses can offer multilingual support more efficiently, enhancing engagement while meeting compliance requirements like Quebec, Canada’s Bill 96.