With Voice-First devices market expected to reach $13 billion by 2024, a number of technology companies are investing heavily to make inroads into this segment. With millions of smart voice devices deployed every month in our homes and businesses, voice is fast gaining impetus as the next primary interface for communication between machines and humans.
Voice devices and channels are similar to how the internet browsers helped us, search information and perform transactions, but much more personalized and smarter ways.
UI/UX designers have driven the adoption of internet for communication and transactions by the masses. Over the years, they have ensured user-friendliness of websites by closely observing users and their behaviour, including non-digital natives. UI/UX designers have had a deep learning curve, having been pivotal in driving human-computer interactions and developing persona-based interaction mechanisms for apps and websites.
Just like designing UI/UX for websites, one of the toughest parts of designing cognitive conversational Voicebots or Intelligent Assistants is to define their personality. An Intelligent Assistant must recognize a user’s intent, context and dialogue timing to ensure it successfully humanizes the conversation. Additionally, it must have a distinct voice, dialogue personality and possibly humour to humanize it further.
If Siri and Amazon Echo sounds human to you and me, it is because dedicated team of writers and UX specialists worked on multiple languages around the clock. The engaging experience with the Doctor Strange voicebot for Skype is attributed to an Emmy Award-winning writer engaged full-time to model the voicebot’s personality from the Doctor Strange movie script (https://blogs.skype.com/news/2016/11/03/enter-the-world-of-marvels-doctor-strange-with-our-bot-and-emoticons/).
Assigning a personality to a voicebot ensures that your system controls the narrative while resonating with the user. The role of a dialog writer becomes a critical piece in this scenario. Dialogs written for voicebots should be clear and simple to engage a user. A dialog writer makes all the difference to the success of a voicebot. Here are a few things a dialog writer will drive to ensure a distinct, meaningful and engaging personality to an enterprise’s Intelligent Assistant.
Ensuring the voicebot is an extension of your brand:
When marketing managers define a customer persona for their brand, they really try to get into the shoes of the customer and picture his/her demographics, daily chores, hobbies, traits etc. Defining the personality of an Intelligent Assistant is no different. As a dialog writer, you have to understand and mirror the user it is trying to address (unless it's a fun persona like Doctor Strange).
You have to come up with the right tonality and dictionary that can be identified by its users and not overwhelm them.
Determining the Dialog Goals:
Every voicebot dialogue has to be written in a way that it serves the users’ needs in the fastest possible time frame. Similar to UX designers, dialogue writers will need to understand users and the problem they are trying to solve. Contexts can vary from answering simple service questions, assisting in making purchases to complex business intelligence queries - as addressed by ConverSight.ai, a conversational analytics, and insights-driven AI platform.
Developing a Dialog:
As a dialog writer, you will need to design the conversational flowchart. You have to have an overarching view of all possible directions a conversation can take during a session between the voicebot and the user. Since every user request can be replied in one or more ways, it is imperative your flowchart covers all possible scenarios comprehensively.
You will also have to cover multiple scenarios, since each user only experiences one scenario with the voicebot, at any given point in time. Hence, you have to ensure that the dialogue for a specific user journey will not account for the other options, which the specific user doesn't care for.
Apart from the above points, the dialog you develop have to ensure that your Intelligent Assistant directly addresses a single person, converses in short sentences, doesn’t go emotionally overboard and also maintains a step driven approach to offer advice.
As you would have realized, dialogue writing for voicebots is a challenging new role in the new IT environment and is by no means a walk in the park. It requires specialists to bring a UX approach to script writing. In all probability, they will be the right people to streamline and contextualize an infinite number of variations in communication.
If you are keen to know how a conversational AI platform like ThickStat’s ConverSight.ai works from a dialog writer standpoint, write to us at firstname.lastname@example.org.