Main illustration: Olenka Malarecka
Artificial intelligence is radically redefining the customer service landscape. From automated emails to visual search, AI allows companies to better support their customer’s needs at more touchpoints along their journey.
Nowhere is this radical change to the customer experience as apparent as in the new wave of chatbots. AI-powered chatbots have become widely available and let you do what was once impossible: help customers 24/7, answer questions without any human intervention and provide support to multiple customers at once.
“Intercom’s Answer Bot answers roughly 29% of all questions before they make it to a human”
At Intercom, we are deeply embracing automation and bots to help businesses dramatically enhance their customer experience and achieve faster growth. That’s why we built Answer Bot, an intelligent chatbot that automatically and instantly resolves your customers’ most common questions.
Here’s a look at how customer service chatbots can improve your service experience, and two case studies of self-service bots that will inspire you to create your own.
What is a customer service chatbot, and do I need one?
A customer service chatbot is a bot that uses artificial intelligence and machine learning to answer basic customer questions via a live chat messenger. These may be questions like “how do I add more users?” or “what is your pricing?”
AI chatbots use your existing information and resources, like FAQs or knowledge bases, to answer questions and offer help. They can recognize and answer multiple forms of the same question and can be trained to give instant responses using your preferred voice and tone. This all combines to create an experience that models what you’d provide customers in a chat with a real person.
You should think about adding a chatbot feature like this to your website if:
- You have a global, growing base of customers.
- Support is needed across multiple channels (website, in-app, on social media).
- You’re unable to have your support team present 24/7.
- You need to reduce your customer service costs and find more efficient ways to scale your customer support.
Consider implementing live chat or targeted messaging first if you have yet to do so: this will still help reduce your support volume and improve your customers’ experiences no matter the size of your business.
Bots help — but don’t replace — customer service agents
It’s definitely true that chatbots help your customer service agents (we call them customer support specialists here at Intercom) do their jobs better. One of the biggest perks comes in the form of time saved: since chatbots take care of simple questions quickly, your team has more time for the complex ones. In fact, Intercom’s Answer Bot answers roughly 29% of all questions before they make it to a human.
“As advanced as AI has become, it can never really offer a genuine ‘I’m sorry’”
AI chatbots aren’t a replacement for real human interactions, though. Bots are better at augmenting these interactions and are best utilized to simplify tasks and remove repetition from workflows. Human agents should handle conversations where someone is navigating a complex purchase or upgrade or is feeling frustrated or confused. As advanced as natural language processing has become, it can never really offer a genuine “I’m sorry” or make things right like a human can.
For real-time interactions, support bots can really have some major benefits both to your team and your customers. Just remember that AI is a virtual assistant, it is there to help your human agents do their jobs better — it can never replace them entirely.
The five major benefits of customer service chatbots
Using AI-powered chatbots to supplement your support team’s efforts doesn’t just make you look smart: it also comes with some pretty big benefits. Here are the five you’re most likely to notice right away.
Chatbots offer an instant resolution
When a customer asks a chatbot a question, it doesn’t just offer a suggestion: Intercom’s chatbots are set up to directly answer questions using your own source material, including knowledge bases and FAQs.
If your chatbot can’t provide the right answer, it won’t block access to humans: customers can request an agent in a single click and still get the fast, accurate answer they need without turning to other channels for help.
Your business can offer 24/7 support
Chatbots, unlike humans, have no need for sleep. If your team is unavailable, a chatbot can step in to answer questions and provide links to resources. But if they can’t help, the bot can indicate your available hours to say when a human will be in touch.
Alternately, if you find that your bots are getting a lot of complicated questions that almost always require human intervention, you can build a dedicated “out of office” bot to take over when no one is around to respond. This way, the customer isn’t waiting around for a response, and a member of your team can reply directly as soon as they’re back.
Chatbots continuously learn
Chatbots don’t just use AI to answer questions. They also use it to learn and automatically improve the quality of the support offered in the future.
For example, if a chatbot knows how to answer a question like “how do I add another user?” it can automatically recognize “where do I add another user?” as having the same meaning. That similar phrasing can automatically be added to its knowledge bank, so future questions phrased in the same way — like “where do I manage users?” — are answered correctly.
“You can train your bots to provide better answers”
Additionally, if a customer rates an instant response given by a chatbot as unhelpful, it’s less likely to use that answer again in a different conversation. If your chatbot tells someone where to add users when what they’re really asking about is user images, a simple rating can help it learn that these two aren’t the same thing.
Of course, not all the learning has to be done via AI. You can also train your bots to provide better answers. You can use questions your customers have actually asked (or ones you think they will ask) to improve how your bot responds.
By training your customer support bot, you can improve its instant responses and increase the likelihood of an self-service resolution being provided to customers, ensuring that no one is left feeling lost.
Your team has more context on each customer
If an AI chatbot is present and patiently waiting, why not use it? Customers who see chatbots may be more likely to ask questions, rather than silently browsing and remaining an “invisible lead.”
With that first touchpoint from a customer, you have their information in your system. This is a great start: if they contact you again, your agents can use that conversation history to better understand their needs or just to make sure their first question was properly answered.
“This deep context on each customer and their journey can give you some ideas for ways to improve your chatbot’s responses”
With all your customer conversations logged in Intercom (including the AI ones), your team can get a full picture of each customer’s journey, how they’re progressing and how the team can support each lead on their own unique path toward conversion or upgrade. If a prospect is asking a lot of questions about pricing, your team will probably want to handle that prospect differently than someone asking a single question about user roles.
This deep context on each customer and their journey could also give you some ideas for ways to improve your chatbot’s responses. If you see some conversations only being held with your agents, this might mean there are questions that your bot isn’t answering but should be able to. That’s a good opportunity to do some training and add those questions into your database.
A consistent user experience is created
Finally, customer service chatbots can help you establish a consistent, on-brand experience for your customers. Since you’re in control of the voice, tone and language used in your bot’s responses, there won’t be any “we don’t say that” situations.
This consistency is only bolstered by how accurate chatbots are. With AI in control, conversations aren’t missed by accident, support emails don’t go to spam and there’s no copy-pasting or mistyping that could result in the wrong information being sent to a customer.
Examples of awesome customer support chatbots
It’s clear that artificial intelligence can improve your customer service experience, both for your team and your customers. Let’s take a look at how some different businesses are using Intercom’s AI-powered support bots to answer questions, provide information and help more customers at scale.
Cleo is an intelligent assistant for your money that allows you to budget, save and track your spending with ease. With close to 200 inbound chats coming in daily and a lean support team, they wanted to implement Answer Bot to help automatically resolve repetitive questions they were getting like “Which banks do you support?” So far, it’s been a massive success!
“Answer Bot helps our team build a better customer experience. Now, 1 in 5 of all our inbound conversations are resolved by the bot!” – Kieran Doyle, Head of Growth
Later is a marketing platform for Instagram that helps visually plan and schedule Instagram posts. With close to 2,000 inbound conversations happening every week, COO Matt Smith saw an opportunity for Answer Bot to help automate answers to questions they consistently received.
From the very beginning, Answer Bot was able to add value by surfacing their most common questions. Previously, they had tried to do this sort of analysis by tagging conversations (difficult because everyone has to tag in a consistent manner) or just asking the customer support team (subject to recency bias). Matt was immediately excited by the clustering analysis which showed their most frequently asked questions.
“Answer Bot immediately delivered value for us on day one. The question clustering analysis quickly identified our most common questions which has always been a manual, time-consuming process. It was easy to get started, and our hour-long investment in setting up and training Answer bot saved us 10 hours in the first week!” – Matt Smith, COO
Aside from the hours saved by using Answer Bot, one important benefit that Matt called out is how it improves their support team’s morale and productivity. Instead of answering repetitive, mundane questions, the bot enables the team to tackle their customers’ most complex issues. Ultimately, Answer Bot is able to augment the team’s effectiveness by allowing them to focus on the most important issues at hand.
Redefine your support experience with AI-powered chatbots
Artificial Intelligence is changing how we live and work. It’s also changing how we’re able to support our customers, whether they’re asking simple questions or trying to make a purchasing decision.
By adding AI-powered chatbots to your site, you’ll do much more than look tech-savvy (although you will). Using a customer support chatbot will save you time, improve the customer experience and help you build a more robust support experience on your site or in your mobile app.
If you’re ready to get started, build your first customer service bot with Intercom now or check out how we got started. We can’t wait to see what you create!