In remote customer service, the era of voice assistants, or voicebots, is coming.
A room with cubicles where consultants in the flesh answer questions through a phone starts to be replaced by a contact center that is more and more often a computer program handling full customer service.
When discussing voicebots, we mean, among other things, customer service processes, voice technologies, and voice user interfaces (VUIs).
When writing about a voicebot, we must also summon related concepts such as Automatic Speech Recognition, Natural Language Processing, Natural Language Understanding, and Artificial Intelligence (Machine Learning).
We are also dealing with the technologies that have become the standard in many industries (e.g., shipping) and that have, for a long time, stopped being a merely imperfect curiosity.
Bear in mind that the voicebot technologies and solutions also offer process automation, better Customer Experience (CX), huge potential, and the power of artificial intelligence worth investing in and implementing.
Voice assistants, voicebots, and phone bots have been replacing telephone consultants for a long time. They have been used for optimizing processes (e.g., request handling), faster offering up-to-date information (e.g., status information), and more effective user query handling.
In this article, we will also look at the function of voice assistants. We will try to answer the questions: How does a voicebot work? How to design a bot that can carry successful conversations?
We will also discuss the applications of this technology, its advantages, and its limitations. We will also briefly reflect on the process of voicebot design.
Undoubtedly, the design, implementation, development, and optimization of voicebots, also in terms of user and customer experience, is forward-looking. From the purely economic and business point of view, too.
It is estimated that the cost of bot work is several and, in some cases, even several dozen percent lower than human labor.
For this reason alone, it is a technology that will determine a company's competitiveness. Voice interfaces pose a serious challenge to traditional graphical user interfaces.
And for these reasons, it is worth reading our article.
Enjoy the reading!
We Audit. We Research. We Design.
Voicebot — what is it?
The era of conversational interfaces, chatbots, and voicebots is just beginning.
And we, people, users, customers, business partners, and even patients, will increasingly interact with bots instead of listening to hotline consultants.
We will talk to them in mobile and web application chats, speak to the devices as part of the IoT (Internet of Things), or have telephone conversations with them.
As usual, the example comes from the top, and the systems offered by the biggest players in the market are the vanguard of change.
Google created Google Assistant. Amazon offers Amazon Alexa. Microsoft prepared Cortana for its customers, and Apple is developing Siri.
Investing in voice assistants by the major players in the ICT industry is no coincidence.
From the article "Are chatbots and voicebots the future of communication? More and more companies are using them," we can learn about fascinating statistics according to which already, in 2020, the value of the global chatbot and voicebot market was estimated at $440 million.
And according to Grand View Research forecasts, by 2030, this market is expected to reach $4 billion.
It is a potential that cannot be disregarded.
OK, so much by way of introduction.
Before we discuss voice assistant design in more detail and describe this technology's benefits, advantages, and potential, let us begin with something much more fundamental: the definition.
Voicebot or voice assistant is a technology that enables the user to interact with the system (device or application) by voice. The user communicates with a voice persona of a brand or organization using questions or simply a largely structured dialog.
The voice assistant is a technology based on two pillars: AI (Artificial Intelligence) and NLP (Natural Language Processing).
Artificial Intelligence enables the caller to navigate the Interactive Voice Response (IVR) system using voice and natural language.
In many variants, while using a voicebot, the user does not have to listen to the menu or select appropriate options via keyboard. They simply need to answer by voice.
The distinctive feature of voicebot technology is the capability to understand human speech and its meaning through the software and formulate an appropriate, relevant, and helpful voice response.
Which, by the way, sounds more and more natural in terms of pronunciation, speech fluency, and naturalness of language constructs.
Simply put, voicebots sound more and more convincing.
In many instances, the answers are strikingly similar to human speech, and often it is more and more problematic to tell with whom the user is dealing.
The distinctive feature of voicebots is also their communication speed which should be understood as the speed of information assimilation by the recipient.
Speech and spoken dialog are the fastest ways to communicate and process information. They are much faster than the written word or dialog in the instant messenger window.
The users increasingly appreciate and accept the response speed and accuracy.
Even more importantly, a conversation with a voicebot more and more resembles a natural conversation due to the capability to understand context, linguistic subtleties, and different language registers and styles.
A voicebot allows you to simultaneously:
- Materialize the brand tone
- Extend the brand personality and give it a multidimensional character/multifaceted nature
- Give the brand the desired audio, linguistic and cultural value.
Within just several years, voicebot technology has evolved from primitive to increasingly mimetic speech simulators able to imitate human speech and conversation as regards its sound, accent, pronunciation, dynamics, structure, course, aims, and forms.
More colloquially, artificial intelligence mimics human speech in a manner that is increasingly more:
- Natural
- Perfect
- Flawless
- Individual
- Efficient
- Economical
- Emotional; bots no longer sound hopelessly impersonal, technical, and machine-generated; they are also designed to imitate emotional reactions.
However, to be honest, you must admit that although they can imitate them, they are still not able to identify them and respond appropriately to them.
These issues are summed up by an accurate observation formulated in the article "5 Reasons A VoiceBot Benefits Your Customer Contact."
Its authors rightly point out that modern technology provides the possibility to recognize specific problems in freely uttered sentences.
A conversation with a voicebot is no longer a conversation using commands and keywords but resembles a more natural conversation with a human.
The range of the users' questions, problems, concerns, and reactions can be defined and simulated. As a result, AI-based voicebots function satisfactorily and, in some cases, even quite well.
And as the authors add on a positive but largely justified note, the times when voice assistants drove users crazy are now a thing of the past.
Past shortcomings, limitations, ineptitude, and numerous errors have been mainly solved and eliminated.
Voice assistant, which so far has had the reputation as a tool to be avoided and has been disliked by many users, to a large extent for legitimate reasons, is slowly changing its image to a more positive one.
The voicebot reputation dynamically improves.
How does a Voicebot work?
The capabilities of today's voicebots have also significantly expanded. The technology enables you to achieve much more complex goals and operates in a much more multitask manner.
Compared to their beginnings and legacy IVR systems, voicebot capabilities in 2022 are downright impressive. What exactly can a modern voicebot do, and what does it do?
A modern voicebot can be used to:
- Activate/deactivate services
- Authenticate customers/users
- Conduct customer and user research
- Identify problems and issues a customer cannot handle
- Inform about statuses (e.g., activation), processes (e.g., payment processing), stages (e.g., of purchase), forms (e.g., of product delivery and service provision), dates (e.g., of delivery) and ways (e.g., of providing services)
- Inform about costs (e.g., of shipping), prices (e.g., of products), and balances (e.g., of virtual transactions)
- Remind about dates (e.g., of subscription expiration or invoice payment)
- Confirm (e.g., payments or documentation receipt)
- Extend (e.g., payment dates or contracts)
- Check availability and booked dates (e.g., booked appointments at doctors' offices)
- Alert (e.g., about system interruptions), inform (e.g., about planned changes), and instruct (e.g., how to solve a technical problem)
- Receive applications, complaints, and claims.
Voicebot is typically used to answer questions whose handling can be automated due to their repeatability, typicality, processual character, and rationality.
More generally, whenever a question and answer can be described in the language of cause and effect, processes, relationships, conditions, formulas, results, and structures, voicebots can be employed.
Main advantages and typical limitations of Voicebots
There are at least as many reasons for the popularization of voicebot technology as there are businesses that use it.
In each case, business usability will be slightly different, but putting the issue collectively and universally, you can indicate several significant advantages of voice assistants.
Their advantages are most vividly visible when we compare the capabilities of artificial intelligence and humans.
In such juxtaposition, Artificial Intelligence more and more often defeats humans in a pretty spectacular manner.
Unlike their human counterparts, digital voice assistants:
- Can work 24 hours a day, 7 days a week, and 365 days a year
- Are not subject to the restrictions imposed by labor law
- Do not get tired and do not need breaks
- Can access information much faster and in a more reliable manner
- Can handle more queries in a given time unit
- Do not have to be onboarded, trained, and sent to courses
- Have access to large datasets
- Do not get sick, and do not change their reactions, responses, and commitment depending on the mood, emotions, and life and professional situation
- Do not have to learn foreign languages; they can offer their support in a multilingual formula.
All these features lead to concrete results, in particular in terms of:
- Company finance: savings are measurable, quantifiable, and convincing
- Efficiency: organization of work and processes
- Setting short- and long-term goals
- Competitiveness
- Flexibility: optimizing and enhancing voicebot functioning poses less challenge than changing human modes of behavior.
Furthermore, voicebots can bring positive changes to an organization in terms of the following:
- The efficiency of work and handling of processes (e.g., sales, aftersales)
- Customer and user impressions and experience
- Customer loyalty
- Brand image and branding
- Sales and service effectiveness.
Naturally, voicebots are imperfect and have several limitations that cannot be concealed.
The main limitations and disadvantages of these solutions include the following:
- Transitory nature of conversation: from the user's perspective, the dialog has no history and does not leave a material trace you can go back to
- Single thread: a voice conversation is synchronous; chatbots (text messages) allow you to have multi-thread and asynchronous conversations
- Need for planning: a voice conversation consists of a simultaneous exchange of questions and answers; consequently, the user or customer is required to reserve some time to conduct it
- Task orientation: a voice conversation is structured by a scenario that aims to correctly identify a typical problem and offer a typical solution
- Communication hypersensitivity: while imitating and simulating real conversations, voicebots are at the same time much more sensitive to their disadvantages, e.g., silence during the conversation with a bot is much more meaningful, disturbing, and stressful than when a person talks to a person
- Communication insensitivity: voicebots cannot sense the customer's emotions, relieve tensions and vent frustrations or respond to the message resulting from intonation, accentuation, pronunciation, and tone and volume of utterances
- Average or minimum usability in the case of problems that are rare, non-standard, context-determined, and resulting from typical human-related problems (e.g., claim or complaint about the manner of service)
- Peculiar reaction rigidity: voicebots cannot improvise and react in a flexible manner adapted to the situation specifics (e.g., image crisis-related)
- Expensiveness in the case of handling complex, difficult, multifaceted, and multidimensional processes.
How to design a voicebot or a voice assistant?
You need to focus primarily on several issues to create successful conversational UIs and voicebots for the desired user and customer experience.
First, a voicebot must be designed for relevance to the tasks, the role to be performed, and the notions of an organization and its representatives.
This means that you need to think about the following issues:
- Sex
- Age
- Tone and language form
- The social distance that the bot will maintain through language
- Personality.
In the technical and linguistic sense, special attention should be paid to the following:
- Speed of speech
- Volume of speech
- Clarity of speech
- Distinctness and specificity/typicality of the tone of voice
- Accentuation and intonation
- Use of dialects, sociolects, jargon, and formal and colloquial language.
A voicebot should be functionally, linguistically, technically, and substantively efficient.
A voice chatbot should also be designed as a brand ambassador (and not merely a tool for checking shipment statuses and giving specific answers to Call Center questions).
Voicebot persona design should take into account the following:
- The role it will perform for the user
- Functions it will fulfill for the customer
- Its main purpose (e.g., informational, sales)
- The scope of duties, competencies, and resources assigned to it
- The way of offering solutions.
The range of potential roles is wide. And they determine further features to be considered and designed when creating a voicebot.
The most popular roles a voicebot can perform are:
- Receptionist
- Advisor
- Problem Solver
- Clerk
- Coach
- Entertainer
- Usher
- Shopping Assistant
- Planner.
Obviously, the level of role complexity, their overlapping, complementing, and supporting, is determined by the specific nature of a given industry, market, and the goals and capabilities of an organization itself.
To some extent, the role will also condition the personality to be implemented through language and audio features, which voice recordings can support.
A voicebot personality should be consistent with the brand image, positioning, market position, and its role in the customer's life.
Also, remember that a voicebot should be synchronized with the following:
- The culture of its users
- Customer lifestyle
- Age, sex, and social status
- Specific notions about how to perform a given role; each customer group slightly differently imagines and defines what it means to be, e.g., a helpful advisor or good coach
- Specific expectations for a given role.
The elements supporting the creation of a distinct and desirable voicebot personality are the following:
- Name that should sound appropriate to the role, friendly, distinct, human, and encouraging
- Sex: proper selection of sex is particularly important in the strongly gendered industries (the so-called men's or women's products)
- Image created using avatars and images visualizing the bot as a person
- Range of associations and notions stimulated by tone, pitch, temperature, and texture of voice
- Branding created by using colors, logos, and elements that allow you to associate it with specific aesthetics and build connections with the users (e.g., of a hotline).
Additionally, the following leading conversational platforms can help you deploy your bots:
- Dialogflow
- Amazon Lex
- Microsoft Bot Framework.
If you want to dive deeper into the topic of conversational interfaces, then we recommend the following article, "Voicebot and Chatbot Design," and the chapter on how to design a killer conversational app and meaningful conversational UIs.
What is a Voicebot? Voicebot design. Summary
- Voice assistants and voice interfaces pose a serious challenge to traditional graphical user interfaces.
- The popularity of voicebots makes it increasingly common for us to interact with artificial intelligence.
- It is estimated that by 2030 the voicebot market will be worth ca. $4 billion. This means that more and more often, we will hear software capable of processing human speech on the other end of the phone.
- Voicebots have been replacing consultants for a long time. They have been used for savings, optimization of processes, faster offering of up-to-date information, more effective user query handling, and better communication with the user.
- The voice assistant is a technology that enables the user to interact with the system by voice. The user communicates with a brand's audio or voice persona using questions or a largely structured dialog.
- A conversation with a voicebot increasingly resembles a natural conversation with a human.
- The distinctive feature of voicebot technology is the possibility of understanding human speech and its meaning and formulating an appropriate, relevant, and helpful voice response.
- The distinctive feature of voicebots is also their communication speed which should be understood as the speed of information assimilation by the recipient.
- A voice assistant allows you to simultaneously materialize the brand tone, extend the brand personality, and give it the desired audio, linguistic and cultural value.
- Voicebot is typically used to answer questions whose handling can be automated due to their repeatability, typicality, processual character, and rationality.
- The main limitations of voicebots include the transitory nature and single thread of conversation and the need to plan it.
- A voicebot must be designed for relevance to the tasks, the role to be performed, and the notions of an organization and its representatives.
- The most popular roles to be performed by a voice assistant include receptionist, advisor, problem solver, and clerk.
