Putting the Chat Back into Bots

Amanda
QwipIt Stories
Published in
3 min readSep 8, 2016

I test a LOT of bots. Lately, I’ve been asking myself a question… What happened to bots that could hold a conversation? It seems like the trend is leaning towards bots that provide limited information and interaction within the messaging platform, and drive the user to an app or webpage.

https://qwip.in/watch/57c5edb4b9708d4b5b41d93c

Lame.

The conversation between the bot and user is a beautiful (semi-awkward) moment. It should actually be helpful to the user. Or why else use the bot? We can all Google things on our own; with a browser that doesn’t contain some hidden agenda.

But here’s the secret that most bot makers won’t tell you: They don’t understand online conversations.

:O

Shocker. I know. But it’s the truth.

Here’s another secret: QwipIt understands online conversations. No, I’m not just boasting; because, our team has spent countless hours, days, weeks, and even years studying online conversations.

First, people want to feel included in the conversation and not be talked at. To put this into context, I need you to travel back to a classroom with me. For me, I’m imagining my 10th grade geometry teacher, Ms. X. Everyday she’d sit at the front of the class profusely sweating as she leaned over the overheard projector. She’d be using stinky white board markers to draw out triangles and formulas on clear overhead projector sheets. All class long. Blah blah blah. There was no interaction with the class. No critical thinking. No group practice problems. Just blah blah blah.

The moral is, I HATED that class. I hated her projecting her lesson on me. I hated that fact that she’d occasionally (I really do mean occasionally) wake me up from my slumber to ask me to do a math problem. Which I’d do with the whole class oohing and ahhing. Then I’d promptly give her the “normal” (aka smug) teenage smile, and go back to sleep.

Don’t project to people; include them. This applies to bots too.

Don’t project your content on me. Find a way to include me.

Second, conversational tone counts. Big time. The tone of the conversation is just as important as the actual content. For team QwipIt, we put serious thought into the tone. (Not the content.) Because tone is super important.

The tone should be a reflection of the brand voice. It should reflect the characteristics that the brand portrays.

Tone is hard in bots. There are only so many characters on a screen. So be selective about the tone and the words you will use to portray this tone. Remember, that bot conversations should mimic real world conversations. Which means that sometimes the tone changes throughout the conversation.

Lastly, it’s ok to add a few messages before (or after) you deliver the content. Conversations aren’t just.. hey here’s the latest story about dinosaurs. They are more round-aboutish. Like: Oh hey I heard this great story about talking dinosaurs. Do you want to read about it? Great! I hope you think the last part is funny too.

These messages are crucial to keep the flow of the conversation lively and actually sounding like a real conversation. Don’t believe me? How many times does someone normally directly answer your question? (Rarely!) This is because as humans we communicate with fillers. Filler words. Filler space. Filler content.

Now that you have a few pointers, let’s go build CHATbots!

Follow us on Twitter
Like us on Facebook
Subscribe to our Subreddit
Make a qwip

--

--