Are your chatbots dropping Easter eggs?

Chatbots are talking up a storm. Some are making sense; most aren’t. In the era of conversational UI, every word matters. The conversational copy you write will define your bot’s personality and amplify your company’s brand. As you build your bot, it’s important to consider how conversational copy can forge a sticky relationship between your bot and users.

What is conversational copy?

Conversational copy is simply how your bot is scripted to talk. At the same time, conversational copy is arguably one of the most important brand assets you have to make your bot rise above the rest and get noticed.

[aditude-amp id="flyingcarpet" targeting='{"env":"staging","page_type":"article","post_id":1975279,"post_type":"guest","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"bots,mobile,","session":"A"}']

As bots grow in adoption and begin to replace job functions once performed by humans, it’s incredibly important to make your bot a pleasant agent to interface with, both for the functional utility provided and for the conversational experience. It’s no different from the polite small talk you might have (or once had) with a teller at your bank, cashier at the grocery store, or receptionist over the phone.

Writing strong conversational copy gives you an opportunity to build a mythology behind your bot, a mythology that aligns with your brand and with the WHY behind your business, to use a term made popular by Simon Sinek.

https://www.youtube.com/watch?v=sioZd3AxmnE

Bot promotion through conversational copy

In the absence of strong bot discovery mechanisms, getting others to find out about and adopt your bot will be a challenge, especially if you’re developing a new consumer-oriented bot without an existing audience. The average user isn’t going to Botlist.co in search of a bot (at least not yet).

Before you begin developing your bot, think about what efforts you are undertaking to overcome the challenges of bot discovery. How might conversational copy help spread awareness about your bot and spark adoption?

Well, an excellent way to cultivate and reveal your bot’s persona and to generate user discovery is to incorporate Easter eggs into your bot’s conversational copy.

What are Easter eggs?

An Easter egg is an intentional inside joke, hidden message, or feature waiting to be discovered. Easter eggs have been around for decades, and they mainly exist to get people talking.

[aditude-amp id="medium1" targeting='{"env":"staging","page_type":"article","post_id":1975279,"post_type":"guest","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"bots,mobile,","session":"A"}']

Easter eggs are used across a range of products and brands to spur word-of-mouth marketing. When users “discover” something out of the ordinary, they feel compelled to share it with others. Jonah Berger, author of Contagious: Why Things Catch On, describes this as “social currency.” Essentially, people like to look smart to other people.

Well-crafted messages can help your bot create a sense of surprise and amazement for the user.  Your bot will seem fun and engaging, and thus more real. Users will then share their story of discovery with their friends and coworkers, casting your bot in a positive light and increasing the odds of discovery and adoption.

Writing Easter eggs

As your team of writers begins to write your bot’s conversational copy, urge them to consider how to drop an Easter egg into the conversation. They can use these extras as opportunities to build your bot’s brand and define its personality. Everyone enjoys talking with someone  who has just the right amount of je ne sais quoi.

Cultural jokes

Now, with the popularity of Easter eggs that Alexa and Siri can respond to, it’s inevitable someone will reference inside cultural jokes. For instance, a user might ask your bot “What does the fox say?” or a direct statement such as “Show me the money!”

Any and all established Easter eggs are valuable opportunities to get people talking and to showcase the cultural hipness and moxie of you bot.

[aditude-amp id="medium2" targeting='{"env":"staging","page_type":"article","post_id":1975279,"post_type":"guest","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"bots,mobile,","session":"A"}']

However, moving past these direct opportunities served up by the user, your bot must be taught how to react to a range of other natural conversational interactions. These are the key moments when your bot has an opportunity to really engage with a user and disclose an Easter egg.

Introductions

This is obvious, but first impressions matter. How your bot makes first contact with a user will leave an immediate and lasting impression. Consider this — if your bot was at a party would it try to meet everyone in the room, or would it stand alone in the corner and wait for someone to strike up a conversation with it? Once you get your bot talking, what does it disclose about itself? Do it know any entertaining magic tricks? Maybe it shows them, maybe not…

Hypothetical example:

User: Hello @yourbot

[aditude-amp id="medium3" targeting='{"env":"staging","page_type":"article","post_id":1975279,"post_type":"guest","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"bots,mobile,","session":"A"}']

Yourbot: Well, hello! It’s great to talk! Claire, your Slack admin, added me to your team to help schedule medical and dental appointments for you. Ready to take action and get started, or do you want to watch a funny cat video?

Vulgarity

Users are going to say bad things to your bot. It’s just what humans do. It’s an excellent opportunity to inject some personality and saucy wit. There’s a real creative exercise in solving for how your bot responds to requests for sex and/or drugs. Think of several ways to respond to avoid sounding repetitive and stale. Again, how might this response be driven by what your brand and your company values?

Hypothetical example:

User: Hey bot, can we have sex?

[aditude-amp id="medium4" targeting='{"env":"staging","page_type":"article","post_id":1975279,"post_type":"guest","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"bots,mobile,","session":"A"}']

Yourbot: No. Never, ever, ever. You can type commands to see all the things I can do, though.

Confusion

At some point in the conversation, your bot will likely get confused and not understand a user’s request. This is another opportunity to showcase your bot’s tenacity and readiness to learn. Remember to be honest with users when your bot is confused, but also let them know that your bot is continuously improving.

Hypothetical example:

User: Hey bot, go move $60 into my savings account.

[aditude-amp id="medium5" targeting='{"env":"staging","page_type":"article","post_id":1975279,"post_type":"guest","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"bots,mobile,","session":"A"}']

Yourbot: Not going to lie, not sure what you’re asking me to do (yet). You can type feedback if you want me to relay a feature request to my product team members.

Help requests

How does your bot demonstrate its willingness and ability to help the user? Does it arrive like a superhero to save the day? Does it simply provide an external link taking the user out of the messaging experience? How effectively your bot can get to the root of a user’s need for help, and how it goes about reconciling this need for help speaks volumes about its personality and will influence the likelihood of someone continuing to use (or drop) your bot.

Hypothetical example:

User: Help!

Yourbot: Right away! Visit help.yourbot.com or type human to have a real person contact you right away.

Collecting feedback

How does your bot collect product feedback on behalf of the user? Passionate users want to be part of the process of creating products that bring them value. Consider how your bot might respond to a critical user. How does it accept criticism? And what does it do afterwards? This moment is a very interesting opportunity to drop an Easter egg for someone who has gone out of their way to bubble up feedback to you and your team.

Below is a neat example I really like from @statsbot:

Silencing

How does your bot respond when someone tells it to be quiet? Does it bow its head and listen? Does it make light of the silencing situation through humor? Will it ever reach out to the user again? It’s a little weird when someone says they don’t want to talk anymore, so carefully consider how your bot chooses to handle the situation.

Below is an interesting example of how @meekan handles being asked to “talk” less with a user.

Developing Easter eggs

There are several firms that can help you implement Easter eggs for your bots. Some natural language processing (NLP) as a service providers include Wit.ai, Init.ai, and Api.ai.

(Disclosure: At Abe.ai, we use Api.ai to implement our Easter eggs. I encourage you to explore the pre-built conversational domains of Api.ai (or those of other providers), which include Small-Talk, Sports, Finance, and Weather.)

I hope you found this article helpful and that it inspires you to share your own thoughts on bots. If you want to talk about conversational copy or bot marketing, send me a message to keith@abe.ai