Designing a Better Bot, Part 1: Thinking Strategically
The popularity of bots will be very short-lived if they continue to be novelties and headaches.
Bots are quickly gaining popularity as a way for brands to build connections with audiences in fast and efficient ways by inserting themselves in places where people are already communicating.
We’re very early on in the technology life cycle of bots and assistants. Despite the popularity of Amazon’s Alexa and the ubiquity of Apple’s Siri, most people are completely unfamiliar with bots, AI, and conversational interfaces. But a bot is not the same thing as a virtual assistant, and yours might be the first that someone interacts with.
Here’s the problem: At the moment, most bots are novelties at best and painfully useless at worst, rather than valuable experiences. If you aren’t careful, you’re going to perpetuate that.
If you truly believe that this technology has a potential to become more than a toy or a gimmick, you have a responsibility to the medium itself to ensure that your audience has a great experience with your bot. Can you imagine how difficult it will be in the future for you or anyone else to sell bots as worthwhile if each bot right now provides a mediocre or bad experience?
First impressions are everything, and I’m here to help you think about how you can make the best of yours.
Exist for a reason.
First and foremost, your bot needs to fulfill a purpose. This can be anything, so long as it’s relevant to your audience. Your bot can be a source of entertainment, a utility to get something done, an information provider, or something else entirely. Your bot’s ability to solve a need or be something that people want will justify its existence.
That purpose must be something greater than just a marketing goal like “get our name out” or “sell our product.” Take advantage of the engagement inherent to a conversation; this is your chance to use a new medium to build a more genuine relationship with your audience. If you offer them something of value, rather than just demand their patronage, it will foster more loyalty than even the most effective print ad campaign ever could.
More importantly, why is this purpose best fulfilled as a bot? You need to be able to answer “why is this experience better as an automated conversation?” All of the different forms of technology and media at your disposal make up an infinite toolbox; do not be seduced by the allure of emerging technology without first asking “is this the right tool for the job?”
Context is everything.
Everything is a journey; where did the user come from, how did they get here, and where are they going next? You should consider not just the superficial context of their journey, like what button they tapped to open up a chat thread with your bot, but the logistical, emotional, and psychological aspects as well. When they start talking to your bot, are they calm or upset? Do they have time to spare, or do they need something immediately?
Anticipating context is a key part of good product design, and that holds true for bot design as well. The difference here is that you need to account for the fact that language and conversation require a different kind of mental energy than tapping buttons or swiping through menus.
Discoverability is a fixer-upper at the moment.
The idea that bots are a way for companies to bypass existing chains of discovery like app stores and content algorithms is a key selling point and strength of the medium, but it’s also a complete paradox. There’s no equivalent of an app store or centralized search for bots for any platform that they exist on. At this point, it is just as hard to get people talking to your bot as it is to build an effective one.
This is absolutely something you need to be aware of, especially in the context of the user journey. Have a strategy and tactics for driving an audience to your bot.
Be great at one thing, not mediocre at many things.
Your bot will be infinitely more successful if it does one thing very well, rather than multiple things poorly. This is both for your benefit and the user’s; Even a practiced UX designer will initially find bot design unfamiliar and difficult, and each additional feature will exponentially increase the complexity of the bot and the amount of work needed to create it. That complexity will translate into an experience that is confusing, lackluster, or otherwise unsatisfying.
Leave the juggling acts to Alexa and Siri, and provide something that keeps people coming back. (But hey, don’t let me stop you. If you create a fantastic bot that can do it all, I’ll be first in line to try it.)
Don’t be boring.
To help build that connection with your audience, your bot needs to have an engaging personality and be written well. Without that, it will fail to retain users no matter how useful or informative it is, to say nothing of any attempt to be entertaining. You’ll need to understand both the nuances of conversation and the limitations of the medium and be able to work inside of them. Copywriters, storytellers, and social scientists all have new importance in the design process, and UX designers who can work on multidisciplinary teams are more valuable than ever.
If it ain’t broke, don’t fix it.
If something works just fine in a form that isn’t a bot, don’t turn it into one just to keep up with tech trends. This should go without saying, but any web experience that would be slower, more confusing, or less safe via bot should not become a bot. Take, for example, the UPS bot:
I used this same tracking number on the UPS site and found my package in under 10 seconds with three clicks. Why would I ever use a bot if it can’t do the one thing it’s there for?
Additionally, bots replace machines, not humans. Along the spectrum from “machine that facilitates a task” to “being we have a relationship with,” it should be pretty clear that bots live at the former, not the latter. Bots are a more personable form of a digital experience. If that experience is better served by a human, use the bot to augment or facilitate that, not replace it.
Don’t lie to yourself; buying things from a bot is kinda weird.
Just because all of us who are already on the bot hype train think some part of this is faster or easier, it doesn’t mean that the average person will. Case in point, the 70% of U.S. consumers who aren’t comfortable buying things through a chat interface. Paying money through a social media product like Facebook Messenger is weird and intimidating at first blush, to say nothing of the awkward back and forth it takes to buy just about anything from a bot.
These are new design patterns that nobody is familiar with yet. Ease your consumer into it, and while you’re at it, look for ways to do it better. Shane Mac at Assist has a couple of thoughts on the subject, if you’d like further reading.
Is this the right tool from our toolbox?
Jeff Goldblum sums all of this up nicely for me in 1997’s Jurassic Park:
“Your scientists were so preoccupied with whether or not they could, they didn’t stop to think if they should.”
Before you dive headfirst into the world of bots and assistants, stop for a second and make sure you have answers to the kinds of questions that are so obvious, they often go unasked.
In my next article, I’ll take a look at specific UX practices you should employ or avoid to improve your chances of building an audience for not just your bot, but for the messenger bot medium as a whole.
Thanks for reading!
If you liked it (or didn’t), please let me know. I’d love to hear from you.
Psst. I have a bot.
Do you like games? Do you like mysterious dungeons? Do you like text adventure RPGs? Wait, you DO??? You should say hi here. And by “hi” I mean “Hello, Dungeon.”
More thoughts on bots:
The Underlying Ethics of AI and Bot Design [Coming Soon!]
