In our last blog , we explored the value of chatbots—little computer programs that reliably communicate with customers on your behalf, freeing your Marketing and Customer Service teams to focus on higher-level tasks.
Since these bots are surging in popularity, we’ll review what makes a good bot—and the best ways to create your own:
The Muppets and Miss Piggy
Despite her ego, you can’t help but feel drawn to Miss Piggy. Back in 2015, Disney promoted The Muppets TV show by developing a Facebook chatbot for Miss Piggy’s Facebook profile. They even cross-promoted through a message from Kermit, who quipped that he’d finally get an hour of peace if she was busy talking to other people.
This was the perfect opportunity for fans young and old to engage with one of their favorite TV characters, and Disney wrote Miss Piggy’s personality into her responses.
What Disney did right: It wrote the chatbot to resemble Miss Piggy’s brand. When fans chatted with her, they received the same snark and ego they’re used to seeing her with. They also did a great job of promoting it by playing off the love/hate relationship she and Kermit have. Finally, they launched it in time to promote the real money-maker: the TV show.
Amtrak developed its website’s virtual assistant, Julie, back in 2012, and it’s a wonder the company ever got by without her.
Here’s why: They eventually saw an 800% return on investment. How? For starters, Julie was programmed to help travelers find trips and routes, access boarding information, and book trips. This saved the company $1 million in customer service emails—in one year. By answering 5 million questions a year, Julie sparked 30% more revenue per booking and increased bookings by 25%. Plus, when Julie answers a question, she includes options to upgrade on the next trip, opening the door for an upsell.
What Amtrak did right: Amtrak saw a way to simplify the booking process for its website’s 375,000 daily visitors. Instead of relying on customers to figure out the website’s interface, they provided the option for a chatbot to help each visitor, streamlining the entire process.
KLM on Messenger
KLM is the world’s oldest airline and a big player in Europe. A few years ago, it decided to become a big player in social media as well. KLM added a “Send Message” button to its Facebook page and saw messages increase 40%. To handle the influx, they created an AI chatbot, which they trained to answer more than 60,000 questions. To ensure the bot keeps quality high, its responses are checked by an agent, and a quick survey is submitted at the end of each interaction. Because of the AI, the chatbot gets a little smart each time. The result: high customer satisfaction that continues to get better.
What KLM did right: Priming their chatbot for 60,000 questions required work upfront, but the payoff is immense: KLM’s messenger requires fewer interventions from human agents.
Anymail’s Chatbot Picks Up the Slack
Anymail was a two-person startup, so keeping up with sales and customer inquiries absorbed valuable time. But they noticed something interesting: They received the same handful of questions with each inquiry. To keep up with their customers, they wrote detailed articles to answer each of these questions, then programmed a chatbot to respond to questions by offering the articles. They even took it a step further: When site visitors reach a page, the chatbot delivers a list of question and answers people often ask related to that page. It’s working: 60% of their revenue is thanks to the chatbot, and a third of visitors take advantage of the feature.
What Anymail did right: The team worked smarter, not harder. They created detailed answers to the most common questions, and even presented these articles before customers had a chance to ask. This cut down on future inquiries, freeing the team to focus on other aspects of the business.