What is a Chatbot?
What is a chatbot, how do chatbots work, how can they improve the customer service and which chatbot is the best for your business?
A timely example of how chatbots are supporting and interfacing with the population, the U.S. Centers for Disease Control and Prevention (CDC) are leveraging the technology to help it in its fight against coronavirus.
Called Clara, it is helping concerned Americans get answers to questions about the virus, to self-diagnose symptoms and to find out what next steps they should take. Launched on March 22, it is already handling 1 million requests a day and the technology behind it is now being leveraged by nine other health organizations in COVID-19 affected countries
Nevertheless, there is still a lot of confusion around chatbots – what they are, what they’re used for and how they could benefit a business – and those that still don’t know how chatbots work and the benefits they can bring could be losing out.
What is a chatbot?
A bot is a software application that runs an automated task – usually over the internet. Therefore, in its most basic form, a chatbot is a computer program that, without human intervention, is capable of automating a narrow written (think Facebook Messenger) or oral (think Amazon Alexa) conversation with a person.
As well as an application, we should think of chatbots as an interface. They sit between the user and a service, just like tapping on a button in a smartphone app, clicking on a menu with a mouse, or even speaking with a customer care agent to resolve an issue.
Why are chatbots good for?
So, why use chatbots? Because chatbots are a conversational interface, they offer a sense of control and greater ease of use than an app or searching through a website’s pages. It’s a user experience more similar to interacting with a customer service representative.
However, unlike the customer services representatives at a traditional contact center, chatbots don’t work fixed shifts. They represent a customer service and customer information channel that is always available and that can always provide an instant answer while demanding nothing from the user other than to first pose a question.
Why are chatbots popular?
Chatbots are becoming popular because of the growing global penetration of social messaging apps. Facebook Messenger, WeChat, Telegram and WhatsApp have become the communication channel of choice for the majority of consumers globally, whether they’re talking to friends or interacting with brands. An astonishing 8 billion messages a month are exchanged between businesses and customers on Facebook Messenger alone.
These apps provide the perfect environment for chatbot deployment and each of the major players now has a platform and ecosystem dedicated to hosting chatbots within their messaging services – WeChat has supported chatbots since 2013 and Facebook launched its platform in 2016.
At the same time, thanks to increased processing power and hardware affordability, artificial intelligence (AI) in the form of machine learning and natural language processing (NLP) has taken a major step forward. It is now possible to build chatbots that are capable of having a conversation with users rather than simply answering a list of commonly asked questions.
What are the benefits of using chatbots?
The advantages of a customer service channel that is always open, that can conduct multiple conversations with multiple users simultaneously, and that provides instantaneous responses to questions should be clear to see.
From optimal use of time, enhanced customer service and customer knowledge to instant chat, each of a chatbot’s strengths is perfectly aligned with customer care. Juniper Research data estimates that by 2023, chatbots will save companies 27 billion customer service hours, annually.
Their round-the-clock availability eases the burden on contact center agents. They in turn, are freed up to focus their abilities on handling the high-value customer interactions where empathy and emotion are crucial to delivering a differentiated customer experience (CX).
Every time a chatbot interacts with a user, it generates a conversation log. This provides valuable data insights for companies and can also be used by contact center agents to quickly resolve issues where the chatbot alone wasn’t able to help. For many businesses, chatbots are already proving adept at prequalifying and streaming customers to the right agent for the fastest handle times and greatest possibility of first call resolution.
How does a chatbot work?
For most applications a chatbot works by simplifying and standardising a specific element of your customer service or customer experience – one usually being handled by a different channel or process. But it’s crucial to remember, even if they can have a tremendous effect on your business when used correctly, chatbots are not a magic wand. They cannot replace every aspect of a company’s existing CX.
Even with recent AI breakthroughs, a chatbot cannot replace customer service representatives. A bot’s greatest strength is in constantly and consistently executing common tasks.
A person, or representative, has the emotional intelligence and understanding to address the most complex customer issues. Even in an increasingly digital world, it is impossible to imagine customer service without human empathy and understanding at its heart.
Therefore, chatbots are there to enhance people’s abilities and to focus their work in the areas that really matter and that really add value.
There seem to be hundreds of different types of chatbots. But whether they’re helping people find things to do over the weekend, helping sports fans keep track of their favorite teams or helping personnel get to grips with a company’s human resources policy, there are only three types of chatbots operating behind the interface – FAQ; intelligent; or hybrid. Each works differently and each has its own unique benefits.
What is a FAQ chatbot?
An FAQ chatbot’s job is to get the user from point A to point B via a predetermined series of questions and potential answers. The actions of a FAQ chatbot, say for ordering a pizza, can be charted as a decision tree. There are a finite number of decisions to make – size, number of toppings, side orders and the end result, point B, is a pizza is sold.
Therefore, this type of chatbot is not saturated with artificial intelligence (AI). This means that at the development stage, each potential way a customer could ask the same question has to be considered, integrated and linked to a corresponding predefined answer. The structure of the English language means that there can be at least 10 different ways of phrasing the same yes/no question.
No scope for mistakes
For this reason, FAQ chatbots are at their best when operating in a clearly defined domain. This rigidness of focus means that they’re not going to deliver a game-changing level of customer engagement, but it also means that they don’t make mistakes.
All of which makes them the ideal interactive tool for moving a customer along a sales funnel, removing the frustration from getting answers to frequently asked questions – and thus improving contact center call deflection – or completing forms and registering for a service.
What is an intelligent chatbot?
An intelligent chatbot leverages elements of AI, such as machine learning and natural language processing (NLP) and is powered by big data, to understand a wider range of questions and queries and to deliver a more immersive, naturalistic user experience.
An intelligent chatbot can maintain a conversation – think of an interactive tour or museum guide – or can still transport a user from one point to another, but does so when the context and position of both points keep changing – think personal assistant.
Open to interpretation
But in order to maintain a conversation, or to identify individual words in a question that completely change the context of the request, an intelligent chatbot needs to be able to recognize and act upon a much larger vocabulary and number of intents.
If there are 10 ways of phrasing the same simple yes/no question, there can be 200-350 different ways of asking the same open question or signaling the same intent – i.e., that the user wants something.
Why intelligent chatbots need data
Consider how many different ways there are of asking for a weather forecast based on if the user wants the forecast for this evening or next week, if he or she wants it for their current location or the other side of the world and if they construct the request as a straightforward question or type, “do I need an umbrella?”
An intelligent chatbot needs a sizable dataset of existing intentions and responses to work with as a foundation and, although it will learn from its mistakes over time and get better at delivering its intended purpose, it needs to be trained in order to reach an initial level of understanding and capability before it goes live.
What is a hybrid chatbot?
If we consider an FAQ chatbot as acting like a pipe, enabling the flow of requests to a single answer, a hybrid chatbot is more like a funnel. It can leverage the same AI that gives an intelligent chatbot its smarts, but in order to channel users to one of any number of scripted responses. If we return to the concept of an assistant again, its intelligence lies in understanding what command or service the user requires and then diverting the request to the correct channel.
Sometimes the result is a simple yes/no, off/on, if-this-then-that solution. However, that solution could be something complex that requires human intervention or human emotional intelligence and problem-solving. In these situations, it’s the hybrid chatbot’s role to hand the user off to the right customer service expert who continues the interaction.
The best of both worlds
A chatbot is faster than a person at trawling through data, and is capable of holding multiple conversations with multiple people simultaneously and of repeating the same task over and over again without a drop in quality or consistency. However, no matter how well it is trained and how much relevant data it has to work with, a chatbot has to remain in a clearly defined domain of expertise in order to be effective.
A hybrid chatbot is a customer service chatbot. It allows you to leverage each of a chatbot’s and each of a person’s strengths to offer a seamless super-intelligent experience for end users.
Does your business need a chatbot?
Even if viewed simply as a self-service tool, chatbots provide a robust, interactive and easy to use channel for customers who have frequently asked questions to answer. Likewise they can provide valuable extra support to your existing contact center, helping to prequalify contacts so that the right agent with the right information and abilities speaks to the right customer at the right time to deliver a resolution.
If supported via a live chat window on your website they can provide that positive interaction or engagement needed to move a visitor into your sales funnel, and they never get overwhelmed by the number of contacts coming in and can respond simultaneous to any number of customers at any time of the day or night.
So the simple answer is yes. But only if you are looking for a digital means of augmenting and improving your existing services and customer experience, rather than replacing existing channels or functions.
Which is the right chatbot for your business?
A chatbot’s job is to make it easy for a customer and a brand to have a conversation. Therefore it must reflect customer needs and expectations – and its design should be informed by the type of experience that a business is attempting to deliver.
No matter how cool or clever it could be, your chatbot is not going to exist in isolation. You need to understand where in the customer journey it is going to sit and what it is going to do.
What will your chatbot do?
Chatbot use cases need to be clearly defined. What exactly is it going to do and how is it going to do it? They can be classified by four broad chatbot use cases or families. They can provide a specific service; enable a commercial transaction; create or enhance an experience; or provide entertainment in the form of fun and diversion.
With the use case identified you need to decide if objectives can be met with an FAQ chatbot using a decision tree and scripts, or if you’re going to have to build an intelligent chatbot capable of natural language processing (NLP) and machine learning.
This is crucial because if, for example, the goal is to automate the provision of a service, your chatbot won’t need to engage in immersive conversations with users. So building a chatbot supported by artificial intelligence (AI) will only lead to unnecessary complication and huge development costs.
For the same reason, you must conduct an ecosystem audit to understand your existing IT infrastructure and how the chatbot is going to fit within it. In the case of an intelligent chatbot, every potential AI solution will need to be benchmarked against legacy systems to find the one that’s the best fit.
Hello my name is…
Every chatbot needs a clear identity and it has to reflect your brand and your approach to CX. Therefore you need to decide if it is going to use formal or informal English, if it’s going to have a sense of humor, if it will be masculine, feminine or robotic in personality and whether or not it’s going to have some form of visual representation.
Now it’s time to start planning and developing the conversational architecture. Regardless of chatbot type, it will mean considering every potential scenario that it could be expected to handle. That means performing FAQ analysis, web and archival research, and collecting and collating all the data needed to construct conversational paths and to understand the reasons why a customer contacts your customer service.
Once you have this information you can start writing the chatbot’s dialogue and storyboarding the journeys customers are going to take when communicating with it.
Once built, it’s time to test to make sure the chatbot can deliver the desired CX. If it’s a FAQ chatbot following a decision tree, it will mean trying every single variant of every question for every single predetermined response. For an intelligent chatbot it means having rich datasets of potential questions or intention structures and suitable answers and using that data to train the bot. The bot must understand intentions and words that may have different meanings depending on context. And it must be able to leverage its own conversational history to prompt or steer a user.
Always a work in progress
Once an intelligent chatbot goes live you will need to set benchmarks for performance and constantly review data to look for areas for improvement.
An intelligent chatbot will become better at answering common questions but will need help to continue responding to new requests for answers it is capable of retrieving. For the best continued user experience, every update or change to an intelligent chatbot should bring a noticeable improvement in performance.
But, no matter how well designed your FAQ chatbot or how well trained your intelligent bot, their performance should be measured in terms of how useful customers find it. Developing a chatbot is one thing, training and fine-tuning it so that it meets or surpasses user expectations is another entirely.
From developing a use case and a brand vocabulary to design, development training and deployment, Sitel Group can help you realize a chatbot that is firmly in line with both your brand and your customer expectations. To find out more about how we’ve helped other organizations automate and enhance their CX, speak to one of our Digital Solutions experts.