Over the past decade, digital transformation has led to a number of changes in customer service for the transportation industry — some good, some bad.
For many passengers, the ability to manage their accounts or top up their balance on an app or website has made their customer experience easier and more convenient. But in some cases, these transitions can have growing pains. Some transportation agencies have offshored their customer support call centers, leading to friction and frustration when customers need timely help.
More recently, we’ve seen agencies adopt AI chatbots for customer service applications. These tools — enabled by the rise of generative AI and large language models — make it much easier and more efficient for agencies to support their customers anytime, anywhere. But like the technology innovations that came before them, AI chatbots can also have consequences when deployed incorrectly.
Let’s take a look at what makes the difference between a good chatbot and a bad chatbot and how a careful AI strategy can help agencies scale customer support without turning off their passengers.
The Ups and Downs of AI Customer Support
Chatbots have been around in various forms for decades. So, what makes an AI chatbot different?
AI-generated chatbots use artificial intelligence, machine learning and natural language processing to simulate human conversations with customers. For transit agencies, AI-generated chatbots can easily answer the most common questions regarding their services and products.
Cubic’s customer experience research found that riders call customer service most often for the following reasons:
- Questions about a route or arrival time
- Lost transit cards — blocking, replacing and transferring the balance
- Issues with tapping their transit card
- Forgot the log-in credentials for their online account
- Problems with loading funds to their account
An AI chatbot with secure access to the right data can answer these questions in a matter of seconds, reducing the burden on human customer support employees. We should expect to see more AI chatbots in the future: a recent MarketsandMarkets report projected that the conversational AI market will grow from $17 billion in 2025 to nearly $50 billion by 2031.
AI chatbots offer a range of advantages to both the agency and the end user:
Availability and Engagement
Unlike human employees, chatbots are available 24 hours a day, 365 days a year. Customers no longer have to wait to call during business hours; a chatbot can respond almost instantly, even in the middle of the night. That responsiveness can drive engagement and customer loyalty: if a user messages your agency at 3 a.m. or from a different time zone, they’ll still get the answers they need.
Convenience
Chatbots make it easy for users to accomplish tasks from the comfort of their own homes that previously required them to go somewhere in person. For example, in some cities, passengers with disabilities previously had to travel to apply for reduced fare concessions and verify their identity in person. Visiting that physical location could be difficult, leading to some riders missing out on benefits they were entitled to. With AI chatbots, those customers can now ask questions, use assistive technologies like screen readers and text-to-voice, and verify their identity without having to leave home.
Efficiency
For transit agencies, large customer support teams can be cost-prohibitive. Human support employees need to be used where they can have the most impact on the customer experience — addressing sensitive issues and meeting with riders face-to-face. Chatbots can easily handle instances where a human touch isn’t required, freeing up support employees to focus on the work that makes the biggest difference.
Data Collection
When bots are integrated with customer relationship management (CRM) systems, agencies can easily gather data that can help them improve the customer experience over time. Data collection from customer support interactions can be tedious and hectic for human employees. Bots are able to ask customers about their preferences and seamlessly enter the information into the CRM.
User-friendly
In our customer experience research, 64% of respondents said their satisfaction would greatly improve or somewhat improve if an agency offered online chat as a customer service option. Similarly, 65% of respondents between 18–34 said their satisfaction would greatly improve or somewhat improve if chat via social media was offered as an option. People prefer chatting and texting to talking over the phone — and chatbots can help to meet that preference. Of course, there are risks and challenges when it comes to using AI chatbots. While genAI tools are great at delivering the information they’re trained on, they don’t excel at understanding emotions or frustration. They can’t empathize with a customer — satisfied or angry — which means they can come off as insensitive or out of touch, impacting your brand’s reputation in the process
Chatbots also run the risk of delivering incorrect information. An AI chatbot is only as accurate as the data it’s trained on, and these tools can glitch, get caught in loops or even present bias when built on inaccurate or skewed information. Agencies must also ensure that their chatbots are secure to protect company information and the PII of customers.
How Do Agencies Ensure Positive Chatbot Experiences?
Beyond building chatbots on secure, representative, and reliable data, transit agencies can take several key steps to minimize risks and create positive chatbot experiences for their riders.
Know When a Human is Needed.
First, agencies should have clear processes in place that allow human agents to take control of chatbot conversations when necessary. There’s nothing worse than being constantly redirected to the same FAQ page or led in circles by a chatbot; by training the bot on the rules for escalation, you can help avoid these negative experiences.
Give Your Chatbot an Identity.
Second, an agency’s bot should be conversational and trained to sense emotions and sentiments wherever possible. Humans are social — a natural, friendly, conversational chat should leave them feeling better than a long, neutral support article. Train your chatbot to answer questions directly instead of redirecting customers to external articles. Agencies can also train their chatbots on common pain points and customer personas, giving the bot a better understanding of potential scenarios and how to manage different customer emotions.
Leave No Rider Behind.
Finally, agencies should be careful not to over-commit to an AI chatbot if it means other demographics won’t be supported. While chatbots have revolutionized customer service, they aren’t perfect, and other riders may prefer phone support or in-person interactions. Agencies should always maintain a diverse range of customer service options to ensure that every passenger can find the help they need.