The personality of your chatbot is a golden conversion tool. But building its personality is not just a matter of designing scripts and avatars. How these are technically executed are just as important.
Ineffective search engines present a host of problems that often don’t seem imminent. It could be a customer’s frustration in trying infinite iterations of keywords; or the endless scrolling by a casual researcher to surface the perfect resource. However, these challenges are far exacerbated for the enterprise.
Enterprise search is not like ordinary web search. Such consumer search is for everyday information, such as how to buy groceries or where to find a babysitter. Enterprise search, on the other hand, is for mission-critical information. For example, searching in a nuclear power plant or in a corporation’s financial department.
Chatbot personality: a distinctive set or pattern of human-like traits that a chatbot is designed to portray.
John Doe
Tweet
Having a personality renders your chatbot the quality of being an individual—just like how each of us are our own unique sum of characteristics. A chatbot that has a personality that’s aligned to the company’s values, and thereby the values of its like-minded customers, is the foundation to build trust and loyalty among customers.
Chatbot personality informs every touchpoint the chatbot has with its users. Its personality is mainly created in its way of speech and manner of design, and then performed through a range of intelligent tools. In this article, we’ll guide you through each of these steps with a deep dive on how to technically execute your chatbot’s personality.
Laying the groundwork: what is your chatbot’s personality and persona?
Determining your chatbot’s personality is a creative task at heart, albeit a crucial one. Before handing the job of building your chatbot over to the IT guys, put together your team of writers, designers, marketers and even psychologists to nail your chatbot’s narrative.
Here’s a brief overview of the key questions you must consider to lay the groundwork.
The first task is to determine your chatbot’s persona:
Who is your chatbot’s audience? Think about the primary demographic of your brand’s target audience. They could be a quintessential senior business executive with mission-critical asks, or a tech-curious millennial immersed in pop culture. From there, make judgements on their interests, the way they use language, and the values they might hold. Imagine who are the peers they frequently mingle with and have your chatbot mirror this persona.
What is your chatbot’s purpose? Is it designed to resolve enquires, nurture prospects or perform transactions, or enhance operational efficiency? Your users’ needs speak volumes on what they expect a chatbot to do when they click on the chat bubble. Here, determine the ideal balance between fun and functionality to ensure your chatbot’s persona is on par with user expectations.
How does your chatbot represent your brand? It’s important for your organisation’s brand to carry over seamlessly to your chatbot. Consider what your brand and its values are. Perhaps a reliable and trustworthy healthcare services provider, a customer-centric retail bank, or a creative and radical clothing label. Play around with visual and textual cues to communicate these values, especially in a way that helps your chatbot stand out among others.
Source: TAIGER
Considering all the above, here are the next steps in laying the groundwork for your chatbot’s personality:
Give your chatbot name. Think about something more human sounding rather than artificial, such as SmartBot or Chatty for instance. Amazon’s Alexa plays with alliteration, and Bank of America’s chatbot Erica takes inspiration from the last five letters of the company’s name. Go for an intuitive name that resonates with, or reminds your audience of, your brand, products or services.
Visualise your chatbot. Explicitly visualising your chatbot as a human could be very inspirative in helping users approach it as a person. Ideally, it should be a mirror of the general demographic of your target audience. You could even use a photo of a person to represent your chatbot.
Strike the right tone of voice. Your chatbot’s scripts form the main bulk of its personality. Carefully think about what mix of traits your chatbot should have, in a way that aligns with your brand and users. Some useful traits to consider in the mix are professionalism, altruism, extroversion, humour, and even quirkiness.
Carrying out the plan: technically executing your chatbot personality
Once you have a clear idea of your chatbot’s personality makeup, the next step is executing the plan. It’s a critical element that requires a marriage of branding and tech which many organisations struggle with.
Execution is not limited to simply pumping conversational scripts into your chatbot library and designing an appealing avatar. Instead, there are multiple smarter tech-based tactics to better allow a chatbot’s personality to shine.
Here, we describe 5 unique features that can help create a more human-like conversational experience for your chatbot’s personality to emerge.
Build a guided dialogue to better perform conversational scripts.
Give your chatbot the power of human memory.
Gracefully handle misunderstandings.
Invest in small talk.
Constantly update your chatbot with language-led AI tools.
1. Build a guided dialogue to better perform conversational scripts
The trickiest part about conversing with a user is their randomness. If your chatbot is unable to manage the infinite breadth of asks it might encounter, it loses the opportunity to connect and relate. To deal with this, a guided dialogue tool helps to steer the conversation within clear domains the chatbot was designed for.
A guided dialogue is a chain of question-and-answer pairs between the chatbot and user. The chatbot is in the driver’s seat, and so it prompts the user with a series of possible questions to ask within a specific domain. Say option A, B or C.
It already has answers prepared for each prompt the user might pick. So if the user picks option B, it provides the required response and follows up with new options, B1 and B2.
This continues until the user has reached his or her end goal. It creates a seamless conversation flow, as if your chatbot understands its user’s needs and is able to anticipate enquiries perfectly.
Source: TAIGER
A guided dialogue tool puts in place multiple structured pockets to better craft your conversations through the entire conversation journey. Start with a strong first impression like in the example above, where the chatbot clearly expresses its personality and sets expectations. Then, think of a follow up question you anticipate your user to ask, and craft a response for each.
Use each opportunity to weave in unique characteristics of your chatbot that builds on previous interactions. Keep building from there. Once you’re done, your guided dialogue should look like this organised conversational tree below, where each stage of the user journey is powerful and engaging.
Source: TAIGER
2. Give your chatbot the power of human memory
The ability for humans to remember nuggets of information they received, ranging from seconds to decades ago, is a powerful tool in human socialisation. Likewise with chatbots, having this shared “shared memory” within each momentary conversation is a valuable human-like characteristic to inform its personality. To achieve the power of human memory, contextual memory is one such feature to embed within the chatbot.
Contextual memory enables your chatbot to maintain and use the context of the current conversation. This includes keeping track of preceding questions and answers, to avoid a restatement of information that was already provided.
In the example below, we can see the user asking ‘where can I buy an apple’, followed by ‘how much does it cost’. Even with an ambiguous pronoun, ‘it’, the chatbot is able to identify the subject the user is referring to, and return the right answer—30 cents!
Source: TAIGER
For the user, conversations powered with memory provides the means to exit the contextual loop in a user-friendly way. It not only makes the conversation much more natural, but also extremely convenient.
Furthermore, contextual memory also helps your chatbot relate to the context associated to the user, to provide more accurate and personalised answers. For example, if your user is enquiring about purchasing a particular product, the system can take that knowledge
3. Gracefully handle misunderstandings
Being able to reduce friction as far as possible when encountering a complex question your bot can’t handle is critical in maintaining user satisfaction. Your chatbot won’t always have the answer to every question, just like humans are prone to error. Weaving in your chatbot’s personality when handling misunderstandings is a useful way to remind audiences that your chatbot is somewhat human.
To prepare for misunderstandings and errors, consider these three steps:
Explain the misunderstanding tactfully. It is essential to be very transparent, honest, and modest regarding the abilities of the chatbot.
Funnel users back to your goals. Refocus the user towards what the chatbot can do. Using the guided dialogue function or fallback options to offer alternate ways to get help which your chatbot is prepared for.
Give contact information to a real agent for help. Allow your user to be handed over to an agent for help, or even have them leave their contact details for follow up.
Source: TAIGER
Here is an actionable guide on creating a robust chatbot fallback strategy to further cushion your users from an unhappy experience.
One example is a search engine fallback, where unanswered questions are automatically pushed as a search query into an integrated search engine instead. It would then be able to tap into external content repositories—far larger than your chatbot’s library—such as public sources or even internal content management platforms like Sharepoint.
4. Invest in small talk
Injecting the conversation with pockets of small talk is so common in human speak. In fact, it’s what makes conversation natural. While small talk often veers from functional discourse, there’s nothing small about it. Encountering a chatbot that can successfully handle chit-chat helps to foster trust with the user.
A chatbot prepared for small talk conveys the impression that it has been configured well to handle natural conversation—which is creative and random. Emotionally, a chatbot that can engage in small talk also makes the user feel like it’s really listening. All these encourage users to continue engaging with your chatbot, and trust that it will be able to respond to their queries.
Source: TAIGER
When building your chatbot, take some time to train your chatbot to answer various kinds of small talk questions. This could range from words of greeting, questions that test your bot, or even some emotional ones.
To expedite the process, a built-in small talk library comes in very handy as well. In Omnitive Converse, this library contains a set of generic responses for common user inputs, such as:
“Who are you?”
“Who made you?”
“Are you a boy or a girl?”
Use these small talk opportunities to take a slight step away from mission-critical transactional tasks or enquiry provision. Remember to inject each response with a taste of your brand personality, to nurture a more engaging experience with your user.
5. Constantly update your chatbot with language-led AI tools
Your chatbot persona is not static, and it can be changed as per the needs of your customers and company. Randomness rules the human world, and a chatbot designed to be as human as possible needs to be prepared for this. Developing a unique chatbot personality requires iteration, feedback, and maintenance, so you know your chatbot scripts are always hitting the nail on the head.
Various chatbot building platforms today allow administrators to configure a chatbot’s scripts. However, what’s usually lacking is the accuracy and sophistication of the resultant chatbot performance. Keep in mind that many traditional chatbot building approaches are based on statistical Natural Language Processing. It comes mainly from Machine Learning—a stochastic and probabilistic approach that relies on large amounts of data to learn.
Programming a conversational AI chatbot to understand how to identify questions appropriately takes time. The industry’s best practice is to train your chatbot about 50 phrase variations for each question! Multiply this by the number of questions your chatbot needs to learn and the challenge is clear.
Look out for approaches that take a more symbolic approach to chatbot building instead. For example, Omnitive Converse employs linguistic approach to Natural Language Processing. This enables the solution to understand the logic in words, without requiring too much data. Combined with semantic technology, dictionaries, and ontologies this approach better derives meaning and resolves ambiguity.
Source: TAIGER
Omnitive Converse’s teaching tool makes it possible to train your chatbot with much greater ease. Chatbots are built on a massive knowledge base, or ontology, that first helps your chatbot understand the meaning of the concepts falling within its domain.
For administrators, this means that training a chatbot to understand a question requires just 10 or so phrase variations. From there, the chatbot automatically learns each phrase, and can accurately pick up variations falling beyond the initial 10 phrases.
A chatbot that can easily learn a wide variety of user inputs is the foundation of iterative conversational design. Invest in an intelligent authoring tool that can help you continually monitor your chatbot’s KPIs, and fine-tune questions and answers. With that, you’ll find ample room to craft creative answers with a lifelike personality.
A cognitive trigger that helps customers click
Building a rich chatbot personality is an invaluable human-centric approach to blend the two dichotomous worlds of artificial and human together. People can interact with chatbots in many different ways, for better or for worse.
A chatbot’s personality sets a crucial frame to help determine how humans should perceive it, and how they should respond to its messages. When executed right, a chatbot’s personality could just be the perfect psychological trigger to convert a user to a customer.
About TAIGER’s Omnitive Converse
Omnitive Converse is a conversational AI tool with easy to configure conversational modules with a powerful language-led AI core to enable rich exchanges. Specialising in creating humanistic conversations, its deployments range in sectors from arts and management, lifestyle and fitness, to insurance, training and development, and asset management. Today, Omnitive Converse is the face of multiple enterprises and government agencies to help them unlock better customer journeys.
Speak with one of our solutions managers today on how Omnitive can unlock better conversational AI today.
Take chit-chat to conversion: 5 tech features that develop chatbot personality
The personality of your chatbot is a golden conversion tool. But building its personality is not just a matter of designing scripts and avatars. How these are technically executed are just as important.
Ineffective search engines present a host of problems that often don’t seem imminent. It could be a customer’s frustration in trying infinite iterations of keywords; or the endless scrolling by a casual researcher to surface the perfect resource. However, these challenges are far exacerbated for the enterprise.
Enterprise search is not like ordinary web search. Such consumer search is for everyday information, such as how to buy groceries or where to find a babysitter. Enterprise search, on the other hand, is for mission-critical information. For example, searching in a nuclear power plant or in a corporation’s financial department.
Chatbot personality: a distinctive set or pattern of human-like traits that a chatbot is designed to portray.
John Doe
Tweet
Having a personality renders your chatbot the quality of being an individual—just like how each of us are our own unique sum of characteristics. A chatbot that has a personality that’s aligned to the company’s values, and thereby the values of its like-minded customers, is the foundation to build trust and loyalty among customers.
Chatbot personality informs every touchpoint the chatbot has with its users. Its personality is mainly created in its way of speech and manner of design, and then performed through a range of intelligent tools. In this article, we’ll guide you through each of these steps with a deep dive on how to technically execute your chatbot’s personality.
Laying the groundwork: what is your chatbot’s personality and persona?
Determining your chatbot’s personality is a creative task at heart, albeit a crucial one. Before handing the job of building your chatbot over to the IT guys, put together your team of writers, designers, marketers and even psychologists to nail your chatbot’s narrative.
Here’s a brief overview of the key questions you must consider to lay the groundwork.
The first task is to determine your chatbot’s persona:
Considering all the above, here are the next steps in laying the groundwork for your chatbot’s personality:
Carrying out the plan: technically executing your chatbot personality
Once you have a clear idea of your chatbot’s personality makeup, the next step is executing the plan. It’s a critical element that requires a marriage of branding and tech which many organisations struggle with.
Execution is not limited to simply pumping conversational scripts into your chatbot library and designing an appealing avatar. Instead, there are multiple smarter tech-based tactics to better allow a chatbot’s personality to shine.
Here, we describe 5 unique features that can help create a more human-like conversational experience for your chatbot’s personality to emerge.
1. Build a guided dialogue to better perform conversational scripts
The trickiest part about conversing with a user is their randomness. If your chatbot is unable to manage the infinite breadth of asks it might encounter, it loses the opportunity to connect and relate. To deal with this, a guided dialogue tool helps to steer the conversation within clear domains the chatbot was designed for.
A guided dialogue is a chain of question-and-answer pairs between the chatbot and user. The chatbot is in the driver’s seat, and so it prompts the user with a series of possible questions to ask within a specific domain. Say option A, B or C.
It already has answers prepared for each prompt the user might pick. So if the user picks option B, it provides the required response and follows up with new options, B1 and B2.
This continues until the user has reached his or her end goal. It creates a seamless conversation flow, as if your chatbot understands its user’s needs and is able to anticipate enquiries perfectly.
A guided dialogue tool puts in place multiple structured pockets to better craft your conversations through the entire conversation journey. Start with a strong first impression like in the example above, where the chatbot clearly expresses its personality and sets expectations. Then, think of a follow up question you anticipate your user to ask, and craft a response for each.
Use each opportunity to weave in unique characteristics of your chatbot that builds on previous interactions. Keep building from there. Once you’re done, your guided dialogue should look like this organised conversational tree below, where each stage of the user journey is powerful and engaging.
2. Give your chatbot the power of human memory
The ability for humans to remember nuggets of information they received, ranging from seconds to decades ago, is a powerful tool in human socialisation. Likewise with chatbots, having this shared “shared memory” within each momentary conversation is a valuable human-like characteristic to inform its personality. To achieve the power of human memory, contextual memory is one such feature to embed within the chatbot.
Contextual memory enables your chatbot to maintain and use the context of the current conversation. This includes keeping track of preceding questions and answers, to avoid a restatement of information that was already provided.
In the example below, we can see the user asking ‘where can I buy an apple’, followed by ‘how much does it cost’. Even with an ambiguous pronoun, ‘it’, the chatbot is able to identify the subject the user is referring to, and return the right answer—30 cents!
For the user, conversations powered with memory provides the means to exit the contextual loop in a user-friendly way. It not only makes the conversation much more natural, but also extremely convenient.
Furthermore, contextual memory also helps your chatbot relate to the context associated to the user, to provide more accurate and personalised answers. For example, if your user is enquiring about purchasing a particular product, the system can take that knowledge
3. Gracefully handle misunderstandings
Being able to reduce friction as far as possible when encountering a complex question your bot can’t handle is critical in maintaining user satisfaction. Your chatbot won’t always have the answer to every question, just like humans are prone to error. Weaving in your chatbot’s personality when handling misunderstandings is a useful way to remind audiences that your chatbot is somewhat human.
To prepare for misunderstandings and errors, consider these three steps:
Here is an actionable guide on creating a robust chatbot fallback strategy to further cushion your users from an unhappy experience.
One example is a search engine fallback, where unanswered questions are automatically pushed as a search query into an integrated search engine instead. It would then be able to tap into external content repositories—far larger than your chatbot’s library—such as public sources or even internal content management platforms like Sharepoint.
4. Invest in small talk
Injecting the conversation with pockets of small talk is so common in human speak. In fact, it’s what makes conversation natural. While small talk often veers from functional discourse, there’s nothing small about it. Encountering a chatbot that can successfully handle chit-chat helps to foster trust with the user.
A chatbot prepared for small talk conveys the impression that it has been configured well to handle natural conversation—which is creative and random. Emotionally, a chatbot that can engage in small talk also makes the user feel like it’s really listening. All these encourage users to continue engaging with your chatbot, and trust that it will be able to respond to their queries.
When building your chatbot, take some time to train your chatbot to answer various kinds of small talk questions. This could range from words of greeting, questions that test your bot, or even some emotional ones.
To expedite the process, a built-in small talk library comes in very handy as well. In Omnitive Converse, this library contains a set of generic responses for common user inputs, such as:
Use these small talk opportunities to take a slight step away from mission-critical transactional tasks or enquiry provision. Remember to inject each response with a taste of your brand personality, to nurture a more engaging experience with your user.
5. Constantly update your chatbot with language-led AI tools
Your chatbot persona is not static, and it can be changed as per the needs of your customers and company. Randomness rules the human world, and a chatbot designed to be as human as possible needs to be prepared for this. Developing a unique chatbot personality requires iteration, feedback, and maintenance, so you know your chatbot scripts are always hitting the nail on the head.
Various chatbot building platforms today allow administrators to configure a chatbot’s scripts. However, what’s usually lacking is the accuracy and sophistication of the resultant chatbot performance. Keep in mind that many traditional chatbot building approaches are based on statistical Natural Language Processing. It comes mainly from Machine Learning—a stochastic and probabilistic approach that relies on large amounts of data to learn.
Programming a conversational AI chatbot to understand how to identify questions appropriately takes time. The industry’s best practice is to train your chatbot about 50 phrase variations for each question! Multiply this by the number of questions your chatbot needs to learn and the challenge is clear.
Look out for approaches that take a more symbolic approach to chatbot building instead. For example, Omnitive Converse employs linguistic approach to Natural Language Processing. This enables the solution to understand the logic in words, without requiring too much data. Combined with semantic technology, dictionaries, and ontologies this approach better derives meaning and resolves ambiguity.
Omnitive Converse’s teaching tool makes it possible to train your chatbot with much greater ease. Chatbots are built on a massive knowledge base, or ontology, that first helps your chatbot understand the meaning of the concepts falling within its domain.
For administrators, this means that training a chatbot to understand a question requires just 10 or so phrase variations. From there, the chatbot automatically learns each phrase, and can accurately pick up variations falling beyond the initial 10 phrases.
A chatbot that can easily learn a wide variety of user inputs is the foundation of iterative conversational design. Invest in an intelligent authoring tool that can help you continually monitor your chatbot’s KPIs, and fine-tune questions and answers. With that, you’ll find ample room to craft creative answers with a lifelike personality.
A cognitive trigger that helps customers click
Building a rich chatbot personality is an invaluable human-centric approach to blend the two dichotomous worlds of artificial and human together. People can interact with chatbots in many different ways, for better or for worse.
A chatbot’s personality sets a crucial frame to help determine how humans should perceive it, and how they should respond to its messages. When executed right, a chatbot’s personality could just be the perfect psychological trigger to convert a user to a customer.
About TAIGER’s Omnitive Converse
Omnitive Converse is a conversational AI tool with easy to configure conversational modules with a powerful language-led AI core to enable rich exchanges. Specialising in creating humanistic conversations, its deployments range in sectors from arts and management, lifestyle and fitness, to insurance, training and development, and asset management. Today, Omnitive Converse is the face of multiple enterprises and government agencies to help them unlock better customer journeys.
Speak with one of our solutions managers today on how Omnitive can unlock better conversational AI today.
Search
Archives
Taiger’s CEO recently shared his insights on the current and future developments of Generative AI at a fireside chat hosted by ICEX
May 9, 2023TAIGER’s Omnitive IDP Solution Now Capable of Extracting Information from Vietnamese Documents
April 26, 2023Categories
Meta
Calender