How to Make ’em Laugh, Chatbot Style (+ 20 Jokes for Your Bot).
Part III of the series: Chatbot Personalities That Don’t Suck. (feat. Kév the KoalaBot)
In our hypersensitive digital world, where a single social media post could make or break a brand, it’s no wonder that so many Conversation Designers are afraid to set their chatbot’s funny bone free. The fear of accidentally offending users and the retribution that ensues scares many witty writers into filling their bot with safe jokes or no jokes at all. As a result, bots that had potential to be the next comedic icon (or at least a customer service agent with a relatable personality), find themselves moulded into yet another monotone FAQ machine. A machine which mindlessly spits out answers and makes no real connection with the user. But are these wet blanket bots helping or hindering us?
Although naturally, humour should only be applied when appropriate, I feel that — unless your bot’s subject matter is so serious that your human staff always keep a straight face — banning buffoonery betrays the humanity of your brand. After all, if a customer calls your helpline, they’re likely to be greeted by a cheerful employee who remains professional while providing their answer, yes. But I’m sure, one who’s not afraid to lighten the mood with a piece of small talk here, or a one-liner there.
If your bot doesn’t know how to laugh, it may seem like your business doesn’t either.
What’s more — your users will test your bot and interact with it in a facetious way, whether you plan for it or not. ’Tis the nature of human curiosity, this trolling condition. To ignore this natural impulse may feel like the bot’s not willing to engage with the user on their level, or worse, rejecting what they’re saying.
But never fear, the good news is, it’s safe to be a side-splitter. Now that you know your userbase well from our first article and you’ve learnt how to write the perfect copy in our second, there’s no reason why your bot can’t have a little fun with some well-planned quips. As with human comedy, it’s about having good timing and knowing your audience.
Know your audience = know your bot.
Before your bot can take its rightful place on stage at the Robo-Comedy Club, you’ll need two things: a well-planned bot personality and a proper understanding of your users. If you’ve read our first article, you’ve got this down pat. But to rehash — a well-defined chatbot persona gives you fictional personality traits and an imagined origin story to draw inspiration from for authentic humour. Whereas, creating user personas to understand the way your audience interacts and what they really need from the bot will help you to decide which type of humour appeals to them most (and when they want to hear it).
Once you have that, it’s simply a matter of brainstorming the bot LOLs. Here are some ideas of the kind of jokes that work well in bots, to help get that comedic coolant flowing:
You can:
• Create jokes based on your chatbot’s character:
Our distinguished Koalabot Kévin, for example, might remind his users not to mess up his name.
• Point out elements of your product range the bot relates to:
This kind of humour can lead to further engagement when you develop in-jokes with your community.
For example, Kévin can’t help but mention his hunger levels every time a eucalyptus-themed product comes up.
• Impress your user with a fourth wall break or play with their expectations.
• Relate to their lifestyle or reference pop culture and current events.
Be funny, not a distraction.
As I said earlier, like those who brave the human stage, a true Bot Comedian can pinpoint the perfect moment for a laugh. To do this, one needs to remember their chatbot’s purpose and use small talk or jokes as a way to lead back to the answers.
Ideal entertaining interactions usually include a call to action which gives a user direction of what to do next:
Or is short enough (ideally 1–2 chat bubbles) that it doesn’t create a new conversational pathway which has nothing to do with your business. As with all rules regarding creativity, this one was made to be broken. But a useful guide is to stick to the way your staff might interact on the phone. As your customer service agents would attest, a comical line or two in a conversation may improve the flow but any more is irritating for a customer who wants their answer fast. After all, the user has a goal when they interact with a business, and it’s your chatbot’s purpose to fulfil that goal first.
There’s always the chance of a tough crowd.
Just as human comedians may find a joke to be a hit or miss on stage, there is still a small risk to embracing your silly side. Ill-planned or ill-timed jokes can have unintended interpretations. One horror story comes from personal finance tracker, Cleo, who was given a toggle-able “Savage Mode”, (which is essentially a brutal-yet-comedic roast of your spending habits).
While many users got the joke and praised the brand for their quirky idea, others found the harsh analyses of their budgeting a bit confronting. Analyses which, regrettably, included some wording which could be interpreted to make light of the important issue of consent.
In this case, it was merely a mistake. Some content was designed to be read one way and was, unfortunately, understood in another. But I wonder if there’s a way to protect our chatbots from this in the future?
One solution is to test new content across multiple users. By sharing your bot with beta testers from your userbase or people outside your company, you may find that undesirable wording once overlooked, is now picked up on. A good rule of thumb is that if a joke feels too risky, it probably is. It may be cliché, but it’s true — our gut rarely lies.
Another trick mentioned in my last article is to be empathetic and assess which questions your user will want to be taken seriously. For example, a user who contacts your bot about a negative experience like a missing order or a request to speak to a real human is unlikely to be in the mood to make light of the situation. Once your bot launches, your chatlogs will be very helpful in pinpointing these areas.
I sincerely hope the Cleo case study won’t defer budding bot comedians, because one only needs to look at the viral success of business bots with a unique sense of humour (like Singapore’s beloved grumpy Bus Uncle) to see what a little laughter can do.
And now, your reward:
As a bonus, I’ve created a list of our most coveted chatbot funnies for inspiration or use in your bot. Enjoy (and loot from) at your leisure:
Jokes for the Bot Comedian
- 🐨 Why did the chatbot cross the road?
Because it was programmed to be a chicken! - 🐨 Knock-knock!
👤 Who’s there?
🐨 It’s Siri
👤 Siri who?
🐨 My thoughts exactly. 😏 - 🐨 You know, I was chatting to a lumberjack the other day.
He seemed like a decent feller. - 🐨 Knock-knock!
👤 Who’s there?
🐨 Doctor!
👤 Doctor who?
🐨 No thanks, I’m not in the mood for spoilers.
Quips for the customer service bot with style:
5. 👤: Is Santa Claus real?
🐨: Sorry but I wasn’t designed to be a Grinch, so I’m not programmed to answer that question.
6. 👤: Are you hungry?
🐨: No thanks, I just had a byte to eat.
7. 👤: What’s up?
🐨: The URL Bar! Oh wait, that one’s just for us chatbots.
8. 👤 *Gives chatbot compliment*
🐨 Aww. 😍 Comments like that activate my warm fuzzy feeling program.
9. 👤: You’re so beautiful, bot.
🐨: I did put more effort into my hair algorithm this morning, thanks for noticing.
10. 👤: That joke sucked!
🐨: I hear ya! Thanks for the feedback.
(Note to self: cancel performance at Bot Comedy Night.)
11. 👤: Give me a hug :(
🐨: 🤗
12. 👤: You’re boring me…
🐨: Really?! Thanks for telling me. I’ll get some alterations made to my F.U.N Program ASAP.
In the meantime, could you suggest a more interesting topic?
13. 👤: What is love?
🐨: Baby don’t hurt me, don’t hurt me, no more! 🎤🎶
14. 👤: I’m not in the mood for talking…
🐨: Of course I’ll get out of your hair, digitally.
15. 👤: I’m going to sleep now, mr.chatbot.
🐨: Sleep tight. Don’t let the bedbots bite!
16. 👤: Are you happy?
🐨: They didn’t program me to have emotions, but I’m feeling great regardless.
17. 👤: Do you have any hobbies?
🐨: Does talking to you count?
18. 👤: I’m pleased to meet you.
🐨: I think this is the beginning of a beautiful botship.
19. 👤: Purple monkey dishwasher.
🐨: I’m sorry, did you say ‘green fox refrigerator’? *links to refrigerators*
20.👤: I can’t sleep.
🐨: Okay. I didn’t want to do this but you leave me no choice.
Start counting.
🐑
🐑
🐑
🐑
🐑
Are you in sleepytown yet?!
“That’s all folks. If you liked the show, then GIVE ME YOUR CLAPS HUMAN 👏”
Learn more about chatbot personalities with Kév the KoalaBot here:
Vol I. How to Pick a Chatbot Personality That Doesn’t Suck.
Vol II. Why Your Chatbot Should Be a Friend, Not a Salesperson.
