Understand the Pros and Cons of Chatbots
Chatbots have become a common sight of late in customer service. Eight out of ten times, my interactions almost always start with a chatbot when I try to reach someone in customer service.
It doesn’t matter whether I am trying to reach someone in banking, insurance, healthcare, retail, or my telco; I end up starting a conversation with a chatbot.
After that, depending on their workflow, I feel elated or dejected by the chatbot’s ability to resolve my queries.
In many cases, I find them useful. However, in some cases, I find them to be super frustrating. What I feel is there is no debate on whether or not to use chatbots as a part of customer service, but there has to be a debate on how to make them work well for your customers.
Take the airline industry, for example. During COVID, I booked tickets to travel with a multi-leg itinerary, which involved multiple carriers. When all the flights were canceled, I had to reschedule them to an unknown time as the airline requested me not to cancel as it might lead to a lot of financial stress for them.
As I am a regular traveler, I decided to do that.
The first point of interaction for me was the chatbot, which was not equipped to handle multi-leg, multi-carrier rescheduling. So, with a lot of effort, I had to move out of chatbot interactions to talk to a human agent to get the rescheduling done. It so happened that I had to reschedule multiple times because of the extended lockdown periods.
On the other hand, some industries have successfully deployed chatbots for the benefit of the customers.
Look at this real-world example.
Bank of America’s Erica, a virtual assistant chatbot, handles millions of transactions per year, assisting customers with transactions, bill payments, and spending insights.
Erica reportedly resolves 90% of queries without human intervention, significantly reducing operational costs.
This brings us to the real debate: when should we say no to chatbots, and when should we say yes?
Let us look at some of the use cases from the real world to see where chatbots have made a lot of progress and where they haven’t.
6 Use Cases Where Chatbots Are a Blessing
1. Banking and Finance
I spoke about Erica, a virtual assistant deployed by Bank of America, in the earlier section. Erica answers questions about transactions, helps users save, and alerts them about potential fraud.
Would you believe that Erica has had over 2 billion interactions, helping 42 million customers since its launch in 2018, resolving 90% of queries without human support?
2. Retail
H&M’s chatbot helps customers check order status, track shipments, and find store locations. It also provides personalized outfit recommendations based on the customers’ preferences and styles.
H&M’s chatbot achieved an 86% engagement rate, with users spending an average of 4 minutes interacting with it.
3. Healthcare
Mayo Clinic’s chatbot schedules appointments and answers basic health-related questions. It provides information on first aid topics, including care instructions for handling a baby’s fever, treating a burn, or performing CPR during emergencies.
Recent studies show that 64% of patients are comfortable using chatbots for non-urgent medical needs.
4. E-Commerce
Sephora’s chatbot helps customers book appointments with a Sephora beauty specialist and offers customers help with making purchasing decisions on their own.
Purchasing decisions based on color matches offered by Sephora boosted their sales by 20%.
5. Telecommunication
Vodafone’s TOBi resolves network issues and helps with SIM activations. TOBi, the customer-facing AI assistant, handles around 1 million conversations on a daily basis and is live in 15+ markets across multiple channels, including App, Web, WhatsApp, Facebook, and Voice.
TOBi handles 70% of customer queries without human intervention.
6. Travel
KLM implemented Artificial Intelligence (AI) for its social media channels, including Messenger, to respond to the growing number of messages. It is trained to answer more than 60000 questions. It shares flight information with customers, such as booking confirmations, check-in notifications, boarding passes, and flight status updates.
This has resulted in a 40% increase in customer interactions, where 15% of online boarding passes are sent via Messenger.
6 Use Cases Where Chatbots Should Be Avoided
1. Emergency Services
During the Hawaii false missile alert (2018), some emergency systems routed users to a chatbot instead of a human, causing panic.
As it is, the missile alert was a combination of human error and communication failure. To top it off, the panicked callers were directed to chatbots, which further increased their panic levels.
Chatbots lack emotional intelligence and crisis response skills.
2. Healthcare and Well-Being
Woebot, an AI therapist, works for general support but fails when users express suicidal thoughts.
In something as critical as mental health and therapy, you should always have the option for the chatbot to direct users to human agents.
Human intervention is critical for crisis management.
3. Legal
A leading law firm tested AI chatbots for drafting contracts but found that 70% of AI-generated contracts contained errors.
Legal language is nuanced and requires expertise.
4. Luxury Brands
Luxury brands like Rolex, Louis Vuitton, or Prada avoid chatbots because 80% of their customers expect human engagement.
After all, I am paying a premium, and I am already considered privileged when I am acquiring these brands. I don’t want to be talking to faceless technology bots when it comes to receiving service.
High-value clients expect personalized attention.
5. B2B Sales
An enterprise software company tested chatbots for lead qualification but saw 50% fewer conversions.
Large deals require strategic conversations.
6. Hospitality
Ritz-Carlton avoids chatbots when handling loyalty guests’ special requests. If you want to provide premium service, you should certainly move away from automation.
Luxury travelers expect personalized concierge service.
It should be a combination of chatbots and human intervention. It cannot be either chatbots or human help.
Instead of fully replacing humans, chatbots should be used to augment human teams. Businesses that get this balance right, like Zappos, KLM, and Bank of America, deliver seamless experiences while avoiding chatbot pitfalls.
When it comes to premium service, most brands tend to move away from automation and ensure human interaction.
For instance, my bank directs all my queries to a human agent by identifying my number. I have never had to go through IVR or self-service options.
Use chatbots wisely, and you’ll provide the best possible experiences to your customers.