Skills Required for Conversation Designing
“Conversation designer” is one role that is certainly trending these days on job portals across the world, yet it’s quite new, so there is not enough awareness among even people working in different fields in technology or even among those who work in artificial intelligence industries regarding what a conversation designer really does. It’s not as new as you might think it is, but still, because there are not enough companies that offer conversational AI services, this role sounds intriguing for many.
“That sounds interesting!” is one of the most common responses I’ve heard when I’d introduce myself as a conversation designer. I don’t know if it “sounds” interesting (guess that’s subjective 😄), but the job certainly is enjoyable!
If you’d like to explore this field, if you’d like to apply for a vacancy for conversation designer, or if you’re an HR executive who’s supposed to find an appropriate candidate for this role and you’re not sure what skills to look for, read further.
Again, this can be subjective, and in some cases, they might want a very narrow set of skills, but in this post, I’ll try to list out the most common set of skills that a conversation designer is expected to have.
Fluency in Concerned Languages: This is the most important skill for a conversation designer. A developer or tester can get away with not being fluent in one or more languages because their job doesn’t need this skill. But can a Java developer not be good in programming in Java? No, that simply can’t be acceptable. Similarly, fluency in English and maybe even one or more languages is a must.
Copy-writing: You could be fluent in a language. Hell, you might even have a PhD in language! But how good are you when it comes to writing? Writing by itself requires a set of skills, I’d say. To be able to write well, not only should you be able to communicate fluently in the language, but you should also know enough to communicate effectively. How to be precise? How to express something in different ways (friendly or professional or casual or highly formal, for example)? How to convey just enough information, not too much or too less, in a given word count? How to keep the user/reader engaged?
Copy-editing: Even the best writers could and would make grammatical mistakes or make typos or use inappropriate punctuation. It happens everyday in the world of writing! But it’d not be considered a good management of resources if they’d have to hire someone else to copy-edit as well, so in most projects, those who do the writing (conversation designing) would be expected to copy-edit the content, too. This is another important skill for a conversation designer. Imagine how bad the users would feel if/when they’d see many silly mistakes — nobody would like that!
Data Collection: In my observation, this is what I have found it to be like: some companies expect their conversation designers to do data collection as well, but most companies know enough to hire resources who would do data collection and data cleaning. This requires some basic knowledge of AI. In other fields, like statistical analysis and data analytics, they’d hire professional data scientists and statisticians to do this job. But in natural language processing, since you’re the language expert, they might expect you to do this job as well. Learning the basics of data science and text analysis would help you much.
Data Analysis: Data analysis is a common task in many roles in artificial intelligence. In natural language processing, as someone who would be considered the language expert in your team, you should be able to analyze the text well, looking at it from different perspectives, to ensure that on the language front, everything works well, as expected. This is not a one-time thing. From the beginning of a project, till the day you’d no longer have to be a part of the project, every now and then, you’d be expected to analyze the language, on both sides (machine and user).
Being Comfortable with MS-Excel and Conversation Design Applications: While writers are mostly comfortable with MS-Word, Google Docs, and maybe Powerpoint and software like WordPress, you need to be able to be comfortable with using MS-Excel, too. Why? MS-Excel is one of the most common and most basic tools used for data analysis, so if you’re expected to do text analysis as well, then you better practice using this application regularly. Also, since you’d be working with engineers and managers, Excel is one platform that they’d all be comfortable with, too. And what’s your choice of Conversation Design software? It is on a conversation design software that you’d be spending most of your time on at work, so get used to commonly used applications.
Knowledge of User Experience: In what platforms would your work be used? How familiar are you with these platforms? In case you’re not familiar with a platform, how quickly can you learn about the features, and tell your team how to make the best use of all the features and how to find workarounds for limitations and restrictions on some of these platforms? When to use images and when not to use images? What are the possible things that can go wrong when you use voice instead of text? The choice of platform and the availability of features is something you’d have little say over. This is something that would be decided by the higher authorities, so you should be prepared to design conversations for any platform.
Convincing and Making Negotiations with The Client: In most cases, clients in the field of conversational AI might not know things. It would be your duty to explain things to them, telling them about the best practices in conversation design, and why, sometimes, even if a client insists that they want things their way, it might not be the most effective or the best option. This is not like developing a website; this is a relatively new field, so clients might not understand right away, so might have to prepare samples or presentations or even have long calls with them. Sometimes, you might also have to convince the clients on behalf of the developers. The ability to convince clients is an important skill for a conversation designer, I’d say. In the end, good user experience is what matters the most.
Basic Knowledge of Technology: Come on! This goes without saying, to work in the field of technology, to work with developers, basic knowledge of technology is required. Being a technology enthusiast would be a bonus! Alright, no kidding, in conversational AI projects, you’d be working closely with developers and machine learning engineers. The technology used can vary from company to company. Also, it’s not just one software that you might be using all the time, so you need to be able to quickly learn to work with different applications. You need not know everything about these applications and algorithms, you need not be a technology geek, but having a basic understanding of the technology you’ll be using everyday would always be helpful. It would just make your job easier.
Being Open-minded: Correct me if I’m wrong, but a lot many writers are often highly opinionated. (Hey, that’s something admirable!) But when it comes to this job, you must be open-minded. You should be open to listening to different people’s perspectives and take into account what the stakeholders want, what your management is willing to offer, what the developers and machine learning engineers can do, how much time, effort, and money your company is willing to spend on data collection and analysis, and most importantly, what the end user has to say. Where there are 10 people, you’ll have 11 opinions, yes, and you must make sure that you’ve looked at your project from different perspectives before you make your decisions on how to present your work.
Being Comfortable with Translation: Only those who have literally worked in content translation would know how hard this can be. You know how often Google translate can get things wrong. You might have to redo most of the content when you;re asked to do conversation designs in two or more languages for the same set of use cases. Yes, it can feel as though you’re the only one in the team who does twice the work. But if you happen to be a language enthusiast, you’d hopefully be able to manage this!
Understanding Your Target Audience: Nobody else in the team needs to understand the target audience or end users more than you have to. In that context, you might want to think like a marketing person. Your client and the stakeholders would be happy if the users of their conversational AI would be satisfied. How to make the job easier for those end users? That must be your primary goal. Big companies invest time and money in understanding end-user personalities, but smaller companies cannot afford to do this, so you’d have to gain this understanding with the resources you have.
On the whole, being a conversation designer is something really cool and fun for those interested in writing and technology, but this can be a highly demanding role as well. Often, you might feel as though you’re trying to understand everyone else’s needs and demands, but nobody in the team understands your concerns. That’s because they don’t understand the struggles that come with managing all these aspects of the job: user experience, text analysis, the pain of editing, the need for consistency, the application of best practices, redoing designs, including the minor and major parts, every time the client wants something changed, convincing the clients and your team members to work according to these best practices, buying time, etc.
