Conversation design can be defined as a design language that takes natural conversation between two people as a model. The more similar the interface is to a conversation between two humans, the easier it is for humans to learn how to use it. In conversation design user audio interface design, interaction design, visual design, motion design, audio design, and user experience copywriting disciplines are combined.
Conversation designers can be compared to architects. They prepare a map of possible user actions and while doing that; they must consider both user needs and technological limitations. They determine the outlines of the conversation with a design scheme that defines the conversation flow and underlying logic, representing user experience. Later software developers and project managers gather together to process this design and actualize the experience.
Now that we have defined conversation design, we can talk about why this concept is important.
What distinguishes an intelligent virtual assistant that does justice to “intelligence” in artificial intelligence from other similar ineffective virtual assistant is the persona. One of the common problems among most of these programs is that almost all of them are identical. This is because software developers rather than real writers design these. Resulting conversations move on a single line -and a useful one, though- and they are far from the natural flow of a creative writer. Many virtual assistants respond to users with “I am sorry. I did not understand you.” when they cannot find an answer to their questions, and this is a good example of a situation that can alienate the user to the process.
It is possible to think of a conversation designer as a screenplay writer. Indeed screenplay writers need to know what kind of people are their characters before writing them. In conversation design system and user, characters serve this purpose.
A good persona must be considered clearly enough to have a unique style and personality, and at the same time must be brief in order to be easily included in the conversation during the writing process. In other words, in the case of a well-designed character, the answer to the “What would this character do/say in this situation?” question should be easy for the designer.
System character represents the front-end of the program that directly contacts the user. For a consistent user experience, it is important to create a clear system character. Otherwise, each designer will work based on their own style, and as a result, there will be disconnectedness and inconsistency in the general user experience.
Google Assistant can be given as an example of a consistent system character. Everything it does (what it writes, says, recommends, etc.) and everywhere it is used (general software and hardware overview and in terms of experience) reflects a consistent character.
Third-party software developers need to create their own characters. The process to create this character typically starts with brainstorming to determine the adjectives that define the main traits of this character (friendly, trustworthy, etc.). A shortlist created in this manner will then transform into a short definition of the character -visuals will generally accompany this definition.
When the conversation is designed correctly, the persona will reveal itself. Various brands spend months creating and developing their digital identity. If they can successfully reflect this identity on a virtual assistant, it will be a pleasant experience for users who are tired of identical programs.
Try to think of a couple of user types that will use your program. It is better to have 2-3 distinct user types -such as a student or a working parent. By means of these distinct characters, you will not make the mistake of designing your intelligent virtual assistant according to your own style and expectations.
In a successful conversation design process, it is important to include copywriters who can work in “narrow spaces”, because just like social media content writing, each word matters in this profession. In the limited area offered to you -think about the screen size- achieving creative results with texts, GIF's and emojis are among the talents that will make a difference in conversation design.
At this point, we can seek an answer to this question: “What will be the next stage?” When desinging a virtual assistant, the conversation designer should consider and design all the steps until the user reaches the final step (and this final step might be many things, such as purchasing, subscription, or connecting to a phone line).
Some users may really need help, while others will do their best to make things harder for the bot. You can think of it as a computer game where you affect the storyline by making your own decisions. There is no linear improvement, and the more flexible your virtual assistant design is, the more attractive you will make it for the user. In other words, if you offer more options to the users, the possibility of their return to this service will increase -just like the users who want to unlock different story endings in a computer game.
We believe that the next occupation pioneered by artificial intelligence and virtual assistants in digital marketing will be the Conversation Strategist.
Writing conversation is, in fact, the second piece of the puzzle. First, you need to decide what you want to write. At this point, uncloaking concepts, seeing how brand activities may change over time, and deciding on the real and measurable tasks of a virtual assistant are important steps that will distinguish the successful work from others.
The person who will do this job will create the main conversation flows, determine user roles and what users can do, and decide how the virtual assistant can reach its goal over the flow in the best way possible. Additionally, after your virtual assistant is online, it is important to monitor feedbacks based on the KPI's of the strategist, in other words, to observe the performance of the virtual assistant for future development. Currently, conversation designers and strategists in the sector are the same individuals; however, in the future, we believe that these two fields will be drawn apart in a way that different talents can develop themselves in either of them.
As we learned what a conversation designer is, it is time to talk about the stages of the conversation design. There are multiple tools to create non-linear texts (like those who write stories for computer games use). But since this is a new field, those tools are not advanced yet. Therefore, we will only talk about the basic steps.
In creative writing, like in many other jobs, it is beneficial to prepare a template to see which path you should take and what you should write. After determining a strategy for your virtual assistant, create the templates of all the flows that you may need to reach your target. Remember that you must have a starting point and also follow multiple paths to reach a certain question. This template will be the “backbone” of your virtual assistant, and when you glance over the flows, you will discover that you will need to discover the additions that will make the conversation sound more natural.
In addition to the template, preparing a flow map is one of the best ways to see how the conversation will flow and how the connection between different points will be formed. Will there be cases when a previously asked question will be asked again, or will the user be guided to a different flow in such cases? As your number of flows increase, your map will become more complex. But in this way, you can present your conversation strategy to your customer more easily, and you can use this map as a guide when you are writing it.
After preparing the template and the flow map, you can start developing the text. At this stage, it is better to start creating certain basic flows in the virtual assistant (you need to state these basic flows on the template), and then, as you continue writing, you need to note the points where the user might be confused or you might need a different answer than the normal flow. These will lead you to alternative flows.
Main flows are the paths user can follow to reach the bot’s purpose. Two or three main flows will be the main part of your test and these will mainly intersect at certain points. You can think of the main flows as the section beginnings of a story.
This might sound simple, but the first impression of your bot is extremely important. Especially if you are drawing users via ads, user interaction must start just after the ad. At the same time, you need to think about how to end the interaction. There may be nothing special you do to end the conversation; you may be directing the user to an employee; you may be giving the contact information. Regardless of what you do, entrance and exit interactions are important for user experience and you need to consider them in planning.
Here, the persona we have mentioned above comes into play. No matter how good your texts are, there will be moments when your bot will fail. A technical error, an unpredicted situation, or -possibly- a user who tries to break your bot. No matter what the reason is, your bot will fall short of the expectations. At this very moment, communication is crucial. The most important virtue of conversation designers is to use more creative wording and to offer another solution to the user when they need to say “Sorry, please try again later”. You should plan how the bot will respond to those situations, what it can do to prevent the user from breaking off the experience, where it will position the user in the flow, and when breaking off happens more than once -and believe us it will- what it should do.
Lastly, we would like to talk about the three points that individuals who will design human-machine conversation should consider.
Start from Humans - Talking is a thousand-year-old activity that has developed and evolved among humans. And it means certain patterns and habits. A conversation with a computer should be designed not to go against these patterns and not to cause bizarre situations. That is to say, computers should fit as much as possible to what humans (users) know and are accustomed to. You can create a smoother experience with this approach.
Adapting to Technical Limitations - What we mean here is to consider the situations that are impossible to arise in human-human conversations. For example, when two people talk, it is not possible for one of them to be disabled due to an operating system error. Or the conversation between people does not have to start with a keyword, for example, "OK Google!" You should consider situations as such and figure out the best approach.
Benefiting from Technical Edges - On the contrary to the paragraph above, computers are more advantageous than humans in many aspects. For example, they never get tired of the same question being asked again and again, or they do not resent you for giving orders. Their answers will not include “gap fillers” like “hmm...”, “oh!". They can find and share information faster. Based on this, you can focus on improving the user experience with these advantages and on creating solutions that will go beyond expectations.
If you want to learn how virtual assistants can contribute to your business and learn more about our solutions, contact us now. Catch the future with us!
Discover MindBehind’s advanced conversational solutions to boost your sales and amplify your support services. Using our easy-to-navigate interface, start building your own sophisticated virtual assistant with zero coding skills required.
Take load off of your customer relations team by integrating MindBehind Live to your Zendesk chat.