A significant disconnect exists between businesses and consumers regarding the use of chatbots. While companies are keen on implementing conversational AI to cut costs and enhance customer experiences, consumers generally find these tools frustrating and inefficient.
Adoption vs. Usage: A Forrester Research survey revealed that 71% of employees across various industries reported their companies were experimenting with or using chatbots. However, only 16% of consumers frequently use them, with over a third avoiding them entirely. The primary issue lies in poor design, which focuses on cost savings rather than customer experience, according to Forrester analyst Max Ball.
Common Issues:
- Overpromising and Underdelivering: Companies often exaggerate the capabilities of their chatbots. For example, banks claim extensive functionalities that are not delivered in practice, leading to unmet expectations.
- Language Barriers: Chatbots struggle with understanding casual language, typos, and common phrases. This gap forces companies to continually update their AI to comprehend new vernacular.
- Redirecting Instead of Resolving: Many chatbots merely direct users to relevant webpages instead of addressing queries directly, causing additional frustration for users who seek immediate answers.
- Multitasking Limitations: Users cannot navigate away from chatbots without losing their place in the conversation, particularly problematic on mobile devices.
- Invasive Initial Interactions: Chatbots often request too much personal information upfront, deterring users who are only seeking simple information.
- Human-Like vs. Robotic Balance: Humanizing chatbots can increase user trust but also set unrealistic expectations, leading to greater frustration when issues remain unresolved.
Improving the Experience: To address these challenges, some companies are refining their chatbots to handle multiple queries and ensuring smooth transitions between AI and human agents.
Future Challenges: Introducing generative AI chatbots poses the risk of inaccurate responses. For example, a generative AI chatbot tested by LivePerson provided incorrect information about baggage fees due to outdated training data. Ensuring the accuracy of AI responses remains a critical challenge.