When designing a chatbot, tone matters. Consumers treat chatbots the way they would a human agent – they expect prompt, accurate answers with an appropriate amount of sympathy. So how do you ensure your chatbot serves your customers well while coming across as sympathetic and conscious of context? Start by making it talk like them.
The tone your chatbot takes should depend on your audience. Your bot should mirror your audience demographic and the care situations they’re likely to find themselves in. Are users contacting support because they’re upset or angry? Make sure the bot is conscious of clues and can respond in a calm but not condescending tone. Will customers be booking an airfare and be excited about embarking on a dream vacation? The bot should be able to join in the celebration. Are users nervous or concerned because an order is missing or they made a mistake in their purchase? Chatbots should appear understanding and quell their concerns.
Imagine contacting your bank’s chatbot about potentially fraudulent charges and the bot continually responds with emojis or slang!
When in doubt, err on the side of conservatism. It’s better to come across as too formal and professional than too loose and casual. Imagine contacting your bank’s chatbot about potentially fraudulent charges and the bot continually responds with emojis or slang. A stressful situation would be made that much worse.
Retailers and restaurants can often get away with a more casual, relaxed, and even playful tone, but be sure your bot is able to turn off the friendly vibes and rise to the occasion when consumers have concerns. Even a pizza delivery bot must deal with missing orders or incorrect charges, and customers will want to feel seen and respected – especially by a chatbot.
Take a cue from the messaging platform Slack: When writing bot dialog, less is more. “Clear, concise and human. If you stick to those things, you can't go far wrong.”
When consumers contact a bot, their experience should feel like they’re talking to a knowledgeable peer. Emulate the voice of customer (VOC), which you can collect from chat histories, or from places they leave online reviews, like TripAdvisor, Yelp, and Amazon. Customer interviews, on-site surveys, and social media are other great places to gather VOC insights.
Your customers should recognize themselves in your brand, so take note of demographic factors like age, gender, and location when creating your bot’s voice. For example, the car rental company Hertz uses a voice that’s laid back but knowledgeable, with the approachability of a friend you’d trust for advice. Alaska Air’s is hopeful and experienced, so travelers feel like they’re in good hands and ready for a smooth trip when booking.
When considering the possible support routes your chatbot will have to take, diagram your conversation flows to make sure your bot has accounted for all of the possible conversations users might lead it down. Lucidchart is a great free online tool to use to get you started, and to help see how far your bot can take an interaction before it’s time to pass the issue to a human agent.
The most important tone is taking the customer’s issue seriously. Your chatbot should be prepared to escalate interactions to an agent with trigger words or situations. There’s nothing more frustrating than a customer feeling like they’re going in circles with a bot because it isn’t equipped to deal with their inquiry. The speed and efficiency of a chatbot partnered with the expertise of a human agent is key to delivering the best customer support.
Ultimately, providing your audience with superior customer care over messaging channels means nailing the voice and tone. If you get this right, customers are more likely to become loyal buyers. Start with strong, relatable care, and commerce will follow.
For more on how chatbots can enhance your brand's customer care strategy, check out Aer Lingus's success story: