Why is it that some chatbots are bland and easily forgotten meanwhile others are enjoyable and in demand? You need to know the answer to this question if you want to build an engaging chatbot that falls into the latter category. Luckily, we compiled a list of great chatbot examples to help you find out why some bots are more popular than others.
Some of them are stand-alone bots, others support a brand.
Let’s see what we can learn from them.
Chatbot use case: Gift Shopping Assistant
Channel: Facebook Messenger
Role: Representing a brand
First on our list of memorable chatbot examples is Ralph by famous toy brand Lego which provides its customers with personalized gift recommendations.
Everyone knows that kids love Lego. Chances are you did as well when you were one. But the problem is that it’s not always easy to decide which exact set you should give to a kid.
Maybe a ninja castle, a pirate island, a police station? Or something else entirely?
It can be difficult, especially when you don’t know the kid that well, but still want to surprise them with a nice present.
And that’s where Ralph, Lego’s bot on Facebook messenger, comes in.
It asks you a few questions about your price range, child’s age, their interests, etc. Then it comes up with a personalized gift suggestion. If you like what you see, you can buy the set immediately.
Ralph is a very simple rule-based bot that doesn’t need Natural Language Processing and AI, yet it has been a great success. According to Sumo, Lego has seen a 6x increase in return on ad spend in the markets where they introduced the chatbot (they used Facebook’s click-to-Messenger ads), and it also helped to reduce their cost per conversion by 31% when compared to other ads.
That’s customer service at it’s best.
Chatbot use case: Virtual Artist + Booking + Store Review + Customer Service
Channel: Facebook Messenger, KIK
Type: Representing a Brand
Next on our list of chatbot examples is Sephora’s bot. This brand offers bots on two different platforms and each of them provides a separate set of services.
- Offers makeup tutorials, how-to videos and product reviews
- Provides a quiz that gathers information and makes suggestions (e.g. age, makeup brand preferences)
- Uses Buttons & Emoji (rule-based)
- Redirects users to the mobile site or Sephora app for purchase
- Offers the Virtual Artists feature that lets you try on different makeup looks using the same technology as the popular Snapchat filters.
- Lets you book an appointment for a makeover in one of Sephora stores. (Uses Natural Language processing).
- Connects you with a customer service representative in real time
Sephora aimed to encourage more people to visit their physical stores. And indeed, thanks to the reservation assistant, the brand has seen an 11% increase in their booking rates. The in-store sales also grew as they’ve seen an average spent of $50 per client who booked their appointment on Facebook.
Chatbot use case: Learning assistant
Channel: Duolingo App
Type: Additional Service
Duolingo, a language learning app, created a variety of chatbots to help their users practice languages. Unlike Lego’s Ralph, they are powered by artificial intelligence and this can react differently to thousands of possible answers.
You know how children find it easier to pick up languages than adults? Well, it’s probably safe to say that this is related to their stage of development since it’s a time of life when you are like a sponge when it comes to absorbing new knowledge and new skills.
However, there’s probably also another factor at play, which is children’s lack of inhibition in regards to practicing the new language.
Kids don’t worry much about their accent, grammar, and vocabulary. They aren’t afraid of saying something completely ridiculous. They just don’t care.
Meanwhile, adults are very self-conscious, and beat themselves up over tiniest mistakes, such as mixing up articles “a” and “the”.
Obviously, this severely impairs their ability to practice a language and slows down their progress.
That is precisely the role of Duolingo chatbots. They allow you to practice your language skills without any pressure. There are no witnesses when you mess up so you feel free to try again and again until you get that particular phrase right.
While, initially, these bots were accessible independently within the app, now they are integrated into the main learning flow.
This is incredibly valuable to language learners because it allows them to build confidence in their abilities before trying to interact with native speakers (which can be intimidating).
Chatbot use case: Customer Service
Channel: Facebook Messenger
Type: Additional Service (Going the Extra Mile)
In 2016, Whole Foods (America’s healthiest supermarket) released its Facebook Messenger chatbot to provide its shoppers with a possibility to browse products as well as find recipes and cooking inspiration.
With each message, the bot offers an array of filters and options for different types of food, including international cuisines. The company made it easy for people to find what they need.
Whole Foods smartly leverages the fact that modern consumers spend, on average, over two hours per day on social media and messaging platforms. Once the user selects enough filters, the bot offers a link to a matching recipe which cleverly takes you to the company’s website.
It’s marketing and customer support in one.
Better yet, the bot allows people to take shortcuts to their favorite recipes by using the messaging favorites – emojis. This allows Whole Foods to simulate human-to-human messaging communication (once again) without the need for artificial intelligence.
For instance, if you send a pizza emoji, the bot answers with a pizza recipe.
Chatbot use case: Guide
Channel: Facebook Messenger
Type: Stand-Alone Service
Here we have another of the Facebook chatbot examples. Software developer and entrepreneur Barbara Ondrisek created Mica the Hipster Cat, a venue recommendation chatbot that grew to 250,000 users in just one year.
“I love fancy coffee shops and restaurants! I spend lots of time hanging out in them, meeting friends or working, and I have my own favorite hipster locations where I know the coffee and vibe is just perfect. So I thought it would be great fun to have a chatbot that shares my love for good coffee and food that can be asked for recommendations worldwide.
This way whenever I want to try a new café or when I travel somewhere, I know, I can ask my trusted companion about her suggestions” explains Barbara.
All you need to get Mica’s advice is send it your location and it will immediately provide local hipster hotspot recommendations.
Mica has a lot of fun features.
For example, if it doesn’t understand your question, it tells you that it didn’t quite get it and asks you for your location… And also sends you a funny cat picture!
Other features include funny cat facts, ability to recognize cats in images, and more.
Chatbot use case: Mental Health & Wellness Assistant
Channel: App – Desktop
Type: Stand-Alone Service
Replika is a mental wellness chatbot that helps people reduce stress, relieve anxiety and combat loneliness, as well as get to know themselves better and be happier.
It has a tragic origin story. In 2015, Eugenia Kuyda, the creator of Replika, lost her best friend Roman Mazurenko who died in a car accident.
A month after his death, she could feel her memory of him fading away, so she went through their chat history as a way to rekindle it. Then, she had an idea. What if she could create a replica of Roman from the digital traces that he left behind?
She used text messages and emails, both between her and Roman and between Roman and his other friends and family, to create a chatbot that sounded like the friend she had lost.
Eugenia then made it public. She noticed that people were interacting with the bot in a way she hadn’t expected: they were sharing things about themselves. So she and a friend of hers started working on a new chatbot that became Replika.
When you register, you are asked about your goals and your daily routine, and then you are prompted to create your personal Replika by giving it a name and choosing its gender.
Over time, as it learns more about you, it will start mimicking you, providing you with someone you can talk to that is a lot like you.
It might seem like a strange idea, but a lot of people worldwide say that talking to someone who is always there for them and never judges them has a positive effect on their lives.
Chatbot use case: Personal Shopping Consultant – Lead Generation
Channel: Landing page
Type: Representing a Brand
BuddyNutrition is an online delivery service that sells personalized vitamin shots and smoothies. In other words, they don’t have the traditional product listing their shoppers can browse. Every smoothie or vitamin shot they sell is tailored to individual physical condition and habits of the user.
Not an easy sell, right?
They need their shoppers to submit a great deal of personal information.
This is where the BuddyNutrition chatbot comes into play.
This rule-based chatbot creates a safe environment and makes the experience pleasant and enjoyable. Throughout the conversation, it cross-references users personal details, physical health condition, fitness habits, and taste preferences to create a vitamin shot or a smoothie.
Chatting to a bot is significantly less annoying than filling out a lengthy form. It also streamlines the sales process without requiring intervention from a human consultant.
It also significantly lowers user frustration as the bottle on the right fills up with each answer letting you know how many questions you might expect.
Simple but incredibly effective.
The list above features very different kinds of chatbot examples.
Yet, all of them manage to capture their audience one way or another.
What is the secret ingredient you ask?
The disappointing answer is that there isn’t ONE.
Like many things in life, the answer here is entirely context dependent.
The main takeaway here is that you don’t need to be a big brand or have a huge budget to create a bot that engages and entertains its users.
Let’s break it down, shall we?
There was a time when people were so intrigued by chatbots that merely having one was enough to get them to engage with it.
Purpose is necessary.
It doesn’t need to be an overly grand and complex purpose but the bot does need a reason to exist; a reason that provides someone, somewhere with some sort of value.
As you have seen from the list, it can be anything from helping people deal with insomnia to letting them interact with their favorite movie characters, learning a language or stopping them from feeling lonely. But your bot has to help with something.
Don’t be afraid to be creative about it.
⭐️ A great bot needs a personality ⭐️
What does that signify exactly?
Does it mean you need to create a full-fledged character with a detailed bio?
Though, you do need to define it to a certain extent.
To be more specific, when we talk, we let the implications of who we are slip into that speech. We do it through the choice of words, expressions or references we use.
Your bot needs a basic behavioral reference structure – is it friendly, funny, sarcastic, compassionate, business-like? Is it a person, an animal, a robot or a fantastic creature? Does it use abbreviations in its speech? Does it use slang words or sophisticated industry-specific vocabulary?
Everything that comes across in speech needs to be considered.
And remember, personality usually accounts for the funniest and coolest parts of bot conversations.
⭐️ A great bot must be functional & easy to use ⭐️
Chatbots have the potential to offer the most user-friendly experience because all humans are aware of conversational rules already.
Nothing needs to be explained. At least it shouldn’t need to be explained.
So, for a bot to engage seamlessly, you need to make sure that those rules are honored.
E.g. a bot should always say “hi” and introduce itself to new visitors rather than jump straight to whatever it is the bot is supposed to be doing. Similarly, there should be no dead ends in the conversation. The user always needs to have an option to continue the conversation, go back or opt out.
⭐️ A great bot has to connect with its audience ⭐️
The last but not least important factor is how the bots purpose, personality, and functionality connect with its intended audience.
It doesn’t matter how well you mastered the three previous steps if they are unrelated to your audience, your bot will fail.
More importantly, this connection is not solely about “Who” it’s also about “When/in what instance.”
Design your bot to your audience as well as the context of the situation it is supposed to solve or manage.
In other words, if your bot is designed to help your customers in time-sensitive moments, don’t let it rumble on with jokes or GIFs. Under normal circumstances, your audience might truly appreciate them but in a hurry, they are nothing but unnecessary friction between them and their goal.
A bot is a bunch of code, but the best chatbots connect with you on a more personal level.
So, whether you are creating a bot to entertain, offer companionship, answer customer questions, qualify leads or book flights, hotels, rental cars and appointments you need to think about to whom it is meant to serve.
The whole trick is to make your customer forget that it’s just a bunch of zeros and ones on the other end.
And remember, when it comes to virtual assistants, the smartest design is not necessarily the most complex or the most expensive.