Writing content for a bot
Conversational UI is a new opportunity for content providers.
An earlier newsletter on the profile of a UX writer got a better-than-expected response from subscribers around the globe. It set me thinking - what were the other “modern” (if that’s the right word) content careers created from the needs of the digital world we live in?
Writing for a chatbot to help it communicate with visitors to your website, and some such, is a new craft; something that I, too, have dabbled in. In tech gobbledygook it is conversational user interface or CUI. In short, a conversational interface is any UI that copies the human way of talking/language.
Here’s how Wikipedia defines it:
Conversational interfaces are platforms that mimic a conversation with a real human. Historically, computers have relied on graphical user interfaces (GUI) such as the user pressing a “back” button to translate the user’s desired action into commands the computer understands. While an effective mechanism of completing computing actions, there is a learning curve for the user associated with GUI. Instead, CUI’s provide opportunity for the user to communicate with the computer in their natural language rather than in a syntax specific commands.
CUI essentially means humans talking to a computer, and the computer responding not in binaries but in the same language.
By now, most of you would have come across examples of CUI. The simplest form is when you visit a website and find a chat widget. You key in a few sentences, and the chatbot, appropriately named, responds.
The more sophisticated versions of CUI are Microsoft’s Cortana and Amazon Echo’s Alexa.
But here’s the thing. Even as we get familiar with CUI in our day to day lives, have you sat back and wondered how the bot (a computer algorithm and not a real person) actually responds to your queries? What exactly goes on at the backend? How does the bot get it right, almost 90% of the time?
That’s where script writing or writing content for a bot comes in.
You would have guessed it by now. To create a bot or a digital voice assistant you will need a qualified developer. The other crucial member of this team, though, is a content provider.
In addition to his writing skills, a chatbot writer needs to be a creative person, must have domain knowledge, and must have an active imagination (I shall explain this a little later). This kind of writer can start off with a salary of anything between US $45,000 and 55,000 a year (market rates).
Click on the banner below to discover an exceptional piece of content that managed to catch my eye in my professional outing.
Like a radio or TV script, you need to write a script for chatbots, too. But before that, you need to take certain factors into consideration. Here they are….
To begin with you need to create a personality for your bot, starting with a name. A website selling baby clothes cannot have a chatbot called, “Black eyed Joe”, who responds to a visitor’s queries with a stern, “This is not the place to buy nappies”, right? A chatbot on a fintech site will obviously respond differently than his counterpart on the baby clothes site.
While writing for a chatbot, the content provider must also keep the brand's “voice” in mind. Is the overall communication tone of the brand jovial or stiff upper lip? Obviously, the dialogues for the chatbot have to be in line.
Domain knowledge is another requirement. After all, someone who writes for a kids clothing line must know about that market and line of business.
Now comes the imagination part that I spoke of earlier in this newsletter. Since a writer cannot really anticipate the kind of questions a visitor may ask, but only make an educated guess, he/she needs to “anticipate” queries in order to write down the predetermined responses. Content providers need to put themselves in the shoes of a visitor and try to figure out, given a certain set of circumstances, the questions (and answers) that the visitor may ask or key in. Tough, isn’t it?
But believe me it can be fun sometimes.
Before writing the script complete with dialogues and the subsets, the content provider, the developer, the marketing and brand managers need to have several brain storming sessions. They need to get a fix on what exactly do they want from the chatbot. Of course, everything, including the script is subject to change as more and more people start using your chatbot.
Then, there’s machine learning and natural language processing technologies that have started eating into the role of a conversational UI content provider, but I shall save that for another day.