The usage scenarios for a conversational interface are usually simple and intuitive. Games come to mind when thinking about mission scenarios, because games are low stakes. But the game interface must be designed to meet the user's expectations so as not to bore the user.
The number-guessing game can be a good example of a conversational interface design that does not require any background knowledge from the user, and thus can be used to explore and test the boundaries of conversational interface design.
2. Create avatar portraits
Consider how to display the functions and images of virtual characters (robots). Maintain a consistent, unique brand presence and personality. Avatars can help you design and write dialogue, so make sure you do it early so it's easier to choose the right wording, grammar, and sentence structure.
The virtual portrait of our guess the numbers game case has the following characteristics:
Optimistic, lively and inspiring.
Attractive, witty advances the game and encourages exploration.
Less formal, simple language that appeals to different ages and groups.
We named it "Number Wizard" to give it more personality and satisfy the user's understanding of the meaning of "magic" and the inherent expectations of the number guessing game.
It's important to note that even if you don't give your avatar any personality traits, users will perceive a persona when communicating with it.
3. Write the dialogue
Write down the different conversational scenarios the user might experience. For the Digital Genie game, we came up with the following conversations as a start to help us understand design thinking and best practices for each conversation.
Path 1: The Pleasure Path
This conversation describes a typical round of conversation where the user guesses three times:
So far so good, but if we stop there and start developing this "pleasure path", the game will become very boring. The user may have guessed 99 times or more in this conversation (the user guessed more than 0 to 100), so we have many opportunities to spice up the game and continue to engage users.
Path 2: The Pleasant Path of Playing Two Wheels
This conversation describes the user playing two rounds in a row, resulting in more guesses:
Take a closer look at how we inject personality traits into this conversation. This one includes more rounds of dialogue than the previous one. Such a design makes the game more unique, while also increasing the development cost to implement these special scenarios.
Path 3: Exploratory Guessing
This dialogue describes the user guessing randomly, and the number wizard uses hints to make the user guess the correct number (this time the answer is 23):
Keep users on the right track. Because some users will deliberately test the boundaries of the system to see how the bot responds. The above dialogue guides the user toward the target result by judging the user's input changes and making a more attractive response.
Path 4: Dialogue fixes outside the game scene
In this path, the user suddenly asked a question unrelated to the game in the process of playing the game: "How long is the Great Wall?", the robot tactfully replied to the user: "Do you want to end this round of the game?" to ask for confirmation and advance Dialogue process.
Path 5: Timed out dialogue fix
In the above dialogue, if the user does not respond for a long time, the robot will take the initiative to make different inquiries according to the length b2b data of time, and finally end the round of dialogue voluntarily.
Path 6: The user guessed the same number three times in a row
In the above dialogue, the user deliberately guessed the number 50 three times in a row, but the robot gave an appropriate reply through judgment. Since this is a game, we can handle edge scenes in interesting ways, including guiding the user away from the current game scene. These edge cases deserve serious consideration, because our target users are those who are prone to "exploring the boundaries of the system". Therefore, we can pay more attention to how to meet their needs. You may find the above dialogue scene very similar to path 3. When developing to implement these conversations, you need to pay attention to these similar scenarios to see if you can optimize your code design to meet these scenarios, while maintaining this diversity.
Path 7: The user actively abandons the game
In this dialogue, the user offered to end the game, and the robot could recognize the user's intention to leave and tell the user the correct answer.