A customer experience guide to chatbots
Chatbots have changed the face of customer experience forever. Reports suggest that by 2020 over 90 per cent of businesses will use one – this guide has everything you need to know about chatbots
Rise of the chatbots: An introduction
Chatbots are everywhere. They’re taking customer experience by storm as every website with a customer-facing service seems to be making use of one.
But just what are they?
Chatbots are computer programs that are able to automatically engage with messages when it receives them. For the most part, they require minimal human input to operate as they are powered by rules and, in some cases, artificial intelligence (AI).
Messaging a chatbot is intended to mimic the experience of messaging a friendly sales assistant – or even a friend.
Chatbots tend to use natural language processing (NLP) which is a branch of computer science that helps computers understand and imitate human language and emotion. In addition, chatbots are versatile and can be programmed to respond to messages in a variety of ways depending on your business needs. For example, they can respond to certain keywords and phrases, and even utilise machine learning (ML) to learn how to answer specific queries over time.
Check out this video for insights into AI and the changing face of customer experience.
In a nutshell, a chatbot is an interface facilitating the completion of tasks via text, and without (or limited) human assistance. In other words, it is a robot that presents itself as a human in a messaging platform – and the more seamless, the better.
Why chatbots have become so popular?
The rising popularity of chatbots isn’t a surprise. It has become an integral aspect of the customer journey for its simplicity, convenience, and lack of significant drawbacks.
- Seamless live chat – Chatbots can respond to customer messages instantaneously, which is vitally important for the customer journey. Anyone will agree that being stuck on hold is a frustrating experience that can diminish your perception of a company’s brand.
Conversely, it wouldn’t be financially viable to have an endless supply of human agents on hand to answer every message. Chatbots instead offer a comparatively inexpensive method of ensuring that customers are serviced as soon as they send a message.
- 24/7 service – In today’s day and age customers expect to be able to engage with a brand at any time and anywhere, making this an ideal justification for the use of chatbots. Chatbots can answer messages at all hours of the day. Chatbots never get tired. Gone are the days of waiting for the next available agent, no matter what time it is – a chatbot can have thousands of simultaneous conversations occurring at the same time.
- Freeing up (human) resources – Chatbots are great for answering a high volume of simple questions that require a precise response. Chatbots don’t need to look up answers like their human counterparts do, so this results in a reduced response time, which is key for the customer journey.
- Personalisation – A chatbot can collect information from your target audience, resulting in a customer experience that is personalised. The information can be obtained through social media, a customer relationship management (CRM) system, browsing habits, and the conversation itself. (provide more detail here)
- Proactive customer interaction – Prior to chatbots, most business only interacted with customers on a passive basis. The business would wait for the customer to send a query and they send a response. With chatbots, businesses gain another avenue for marketing and proactive interaction. Chatbots can strike up conversations with users as they are navigating a webpage, offering deals, web links, and general assistance.
The different types and internal workings of a chatbot
The world of chatbots is huge and ever growing. Chatbots are becoming smarter and more complex as the scope of their utility increases.
- Generative – Generative chatbots rely on a repository of information to formulate their responses. In essence, they are pre-programmed to answer questions based on the context of the query. They can be as complex or as simple as they have been programmed to be.
- Retrieval-based – A chatbot that utilises a retrieval-based framework doesn’t rely on a bank of pre-programmed responses. Instead, this kind of chatbot uses a process called deep learning, where the chatbot learns from its experience how to answer questions.
- AI chatbots - This type of chatbot employs artificial intelligence to deliver customer service. The chatbot analyses a customer’s message according to its set parameters to understand the user’s intent. Virtual assistants tend to fall into this category as they are able to analyse keywords and phrases.
- Rule (sequential) chatbots – This type of chatbot follows a script that has been designed by its makers. It can’t deviate from its programming like an AI chatbot can. They often operate in a customer service capacity and send messages via the messaging platform.
While they’re not as complex as AI chatbots, they can determine if they are unable to solve a customer issue and redirect the customer to a human agent.
- Chatbot marketing – Chatbots aren’t limited to handling customer queries. They can also be used to engage customers prior to a transaction and post-transaction. This type of bot can ask if the customer is satisfied with their purchase, relay an offer, and announce new products and services.
- Chatbot analytics – Chatbots can support analytics that analyses data and historical conversations in real time, giving the chatbot a wider pool of information to draw from. Chatbot analytics will help determine problem questions, bottlenecks, and more.
- Social media chatbots – Social media chatbots open the doors to long lasting relationships with customers. Through personalised engagement and marketing. Not only are they able to monitor a customer’s posts on social media, but they can also formulate suggestions based on social media profile criteria.
Chatbots are becoming the new front line customer experience. They are getting smarter and more useful for businesses, but there are a wide array of challenges that chatbot developers and practitioners face.
- Automation + human hybrid – While chatbots are at the forefront of customer-first solutions, they do have many downsides, such as the inability to respond to complex queries. That’s where humans come in. In a collaborative effort, the chat can switch dynamically between human and chatbots depending on the scenario, and the customer is none the wiser.
The chatbots can handle simple queries and take a step back when things get more complex. But, while the human takes over, the chatbots work to provide assistance in the form of relevant customer and query data. This system mitigates the drawbacks of agents that are completely human or bot-based.
- Cost – While chatbots are inexpensive in comparison to human agents, the initial installation can be a drain on manpower and finances. Each bot has to be programmed differently to align with a different business model and brand, while maintenance fees will add to the overall cost.
- Failure to understand a query – Complex questions and new questions will stump a chatbot; this is especially true for retrieval-based chatbots. This can lead to the customer rephrasing the question in vain, long wait times, and customer dissatisfaction.
- Replacing human workers vs enhancing workers –This is more of a cultural issue, but there is no getting away from it. Any discussion regarding chatbots has to mention the cost of human jobs.
That said, chatbots aren’t complex enough to entirely replace human agents. For example, a chatbot wouldn’t be able to deal with a complaint; and probably lead to more complaints. In addition, many people simply prefer interaction with another human.
It doesn't take long to find a chatbot success story. Here are 3 examples of a chatbot done right.
- Bank of America – In 2018, Bank of America launched Erica. This chatbot helps users check their balance, remind users about their bills, and answer a variety of banking related questions.
- H&M chatbot – In 2017, H&M launched a chatbot that was tailored towards providing a personal shopping experience. It achieves this by asking users a series of fashion related questions, bringing gamification into the mix. Users simply pick answers from photo options, resulting in a personal style profile.
- Amatak – Manually booking rail travel can be a huge hassle. Amtrak provides a simple solution let users state where they’d like to go, and the bot fills out all of the forms and does all of the heavy lifting.
A future where chatbots rule
The future of chatbots is very bright. As more data becomes available chatbots will become smarter and their ability to understand complex questions will grow.
- Artificial intelligence – In the next few years, chatbot intelligence is expected to increase substantially. In the future, chatbots that use machine learning, natural language processing, will give them the ability to offer first-pass responses that sound natural. If 2017 was the year of the chatbots, 2018 will likely be the year where chatbots become human-like.
- Business intelligence – As data grows and the ability to process and analyse becomes more importance, the market for business intelligence chatbots will grow as a result. Bots are the gateway to quick and in-depth analytics.
- Deeper customer insights – Chatbots are great at storing data which can be obtained for analysis. Chatbot capabilities in this area are expected to increase substantially, as chatbots will aggregate more data and reliably predict customer behaviour.
- Website and social media chatbots – This is already starting to happen. Businesses the world over will embed chatbots onto their website as an inexpensive avenue towards good customer service. The rapid adoption of messaging apps has also been beneficial for the chatbots market, as they are being deployed on the likes of Facebook, Twitter, Slack, WhatsApp and more. The business potential is huge, and we will likely see even more social media focus in the future.
Key figures and trends
• The global chatbots market is expected to reach $1.23 billion by 2025 (Grand View Research)
• 48 per cent of consumers prefer a chatbot that resolves queries and issues over chatbots that have a ‘personality’ (Business Insider)
• 47 per cent of consumers are open to buying items from a chatbot. (HubSpot)
• 80 per cent of businesses want to have a functional chatbot by 2020 (Business Insider)
• 61 per cent of businesses owners believe chatbots will not replace human jobs (Chatbots Journal)
For more insight into the future of customer experience, check out this video.
Want to know more about customer experience and chatbots? Check out the articles below!
- AI 2020: The future of customer experience
- 5 companies transforming their customer experience with AI
- ‘The greatest challenge in implementing AI is mindset and cultural change’
- Consumer appetite for chatbot experiences on the rise
- What role will AI play in the future of customer experience?
- How Machine Learning helps AI-driven speech recognition
- Defining the new normal with AI: How the Nordics are transforming CX
- Customer experience industry poised to join the AI revolution
- Solving the live agent vs AI conundrum: The human-in-the-loop approach
- Coca-Cola to launch vending machine powered by AI for smarter CX
- Unleashing CX: The power of emotion in experience design
- ‘Customer experience is not only customer service, it’s a big strategic tool
- The positive economics of customer engagement’
- Customer experience management: The 5 competencies of CX success