Talk to the customer, and you’ll find that customer service chatbots are very hit or miss.
This is due to the fact that even though over half of consumers want to use them, many chatbots are still missing the mark. We want the ease of automation. But we don’t want automation if it means a lack of information.
So, what’s the best way to design and deploy your customer service chatbot? Keeping reading!
Chatbots = Convenience
I truly believe chatbots are the easiest way to enter the world of AI. If artificial intelligence seems “scary” or if you think your business can’t take advantage, think again.
A survey by Usabilla found 35% of customers report the number one reason they would use a chatbot would be to save time.
When it comes to customer service chatbots, that means you should consider the convenience factor when designing and planning your chatbot.
Areas to consider are:
- Answering Frequently Asked Questions (FAQs)
- Scheduling appointments
- Updating billing or payment info
- Checking balance or transaction info (finance)
- Checking/changing flight info (travel)
Furthermore, the survey also found that nearly three out of four (70%) of respondents said that they have used a chatbot already. And for those who haven’t, 60% said that they would feel comfortable using a chatbot.
What Will My Chatbot DO?
If you’re feeling better about people actually using your chatbot (they will), you’re ready to move on to DOING the darn thing. As in, what should our chatbot DO?
The first thing to understand about your customer service chatbot is that you must tie its results to business outcomes. (which is really what we say about ALL of your marketing tactics!)
In talking with many of our Customer Care clients, we found that common but high-volume FAQs are usually what bogs their customer service teams down. This, then, prevents them from answering more complex — or as we call them, “red level” — issues.
So for many of our customers, a chatbot is a convenient way to answer common, high-volume, or “green-level” issues that many customers have. This, in turn, frees up the human team to focus on yellow-level or red-level issues that require a human touch.
The business outcome of the above is that human capital is saved (time saved = money saved) while the chatbot is still able to serve the customers’ needs.
And it’s worth noting: A chatbot can answer “green-level” FAQs 24/7 with 100% accuracy.
Take A Human-Centered Approach
Perhaps that sounds like an oxymoron, but it’s not! Here at B Squared Media, we call it, “human-centered AI.”
In fact, we believe that the framework for AI goes something like this: HUMAN —> AI —> HUMAN.
In other words, humans are needed to program, plan, and deploy the customer service chatbots, and then humans take the data the chatbots collect to make iterations and deliver more data-driven decisions.
The bottom line, we keep a human at the center of our intelligent tools loop. And the research backs this up.
- Be transparent (let the customer know it’s a bot)
- Add a personality (maybe even give your bot a name. Ours is DudeBot)
- Brand the thing! (including tone and voice)
- Know your trigger words
- Plan out your conversational flow
Consequently, your conversational flow and trigger words will change over time. As the machine interacts more and more with your customers, it will be able to tell you the paths they’re taking and the words they’re using. This is where humans come back in. Take that data and make a better mousetrap!
Above all, while consumers are willing to use your customer service chatbot, they want immediate help from a human if things are unclear. AI is NOT replacing customer service, it’s merely enhancing it (when done correctly).
Change Customer Service To Customer Care Or CX
Another win for your customer service chatbot speaks to your culture. Customer service really encompasses reactive tasks; as we are called upon, we answer or resolve issues.
Meanwhile, customer care or customer experience (CX) offers a proactive way to engage with your customers. And this is important.
A PwC report defines the importance of customer experience:
While many companies focus significant time and money on design that pops or cutting-edge technology to wow customers, these aren’t as essential to the experience equation as many companies believe. Customers expect technology to always work and often don’t take notice of it (unless it’s malfunctioning). They want the design of websites and mobile apps to be elegant and user-friendly; they want automation to ease experience. But these advances don’t matter much if speed, convenience and the right information are lacking.
When customers’ expectations are met or exceeded, companies gain measurable business benefits—including the chance to win more of their customers’ spending dollars.
The point of AI and human-like technology isn’t to push us further apart. If done well, it should bring us closer together — as well us help us communicate more effectively!
The Future Of Customer Service Chatbots
Chatbots function based on keywords and keyword phrases used by your customers. Trigger words, as I mentioned above, help alert the humans monitoring your chatbot when they need to step in and help.
By periodically reviewing chat transcripts your customers have with your bots, you can help make iterations to your chat services.
- What words and phrases are used most often?
- How well are customers are conversing with our chatbot?
- Are we providing the right solutions at the right time?
- What’s “new” in the interaction log that we need to add to our chatbot’s conversational flow?
AI — and chatbot — technology is ever-evolving. (voice chatbots, anyone?!)
Like any piece of technology or tool, a chatbot will easily fail without proper planning. But if you aim to deploy a customer care chatbot, your human-centered AI can become an invaluable part of the team.
Latest posts by Brooke B. Sellas (see all)
- Managing Customer Care Crisis Communications Through Social Media - April 8, 2020
- 8 Ways To Quickly Up Your Customer Support Game - February 5, 2020
- How To Think Conversation & Collect Customer Feedback [With Examples] - January 8, 2020