For many brands using chatbots, conversations are the hardest thing to carry out. They set up creative welcome sequences, but get little traction. They revamp their pitch, but get no leads. They send funny GIF’s with content offers and discounts, but get little response.
These tactics worked well in a time when messenger marketing channels were less crowded and social media still provided organic reach. But today, there’s too much competition in the consumer’s chat window.
Chatbots have become more personal, and continue to feel and act more human. Audiences are adopting to messenger apps as the main communication channel.
The chatbot ecosystem has changed, but the goal remains the same: quick, creative, and convenient service.
Today, generic broadcasts to your subscriber list don’t always work. In a noisy and competitive space, the best chatbots are novel and act as the direct line between customer problems and solutions.
In this post, we talk about the best chatbot examples, actionable insights you can take from them, and how to create your own today.
Chatbot Example #1: Emirates Vacations
Average CTR for display ads are at an all-time low of .35%. Considering this, Emirates Vacations created a chatbot within their display ads. The company targets different visuals and chatbot sequences based on the page someone’s browsing. Engagement rates rose 87% since deployment in 2018.
Lessons learned from Emirates Display Ad chatbot:
- Be relevant to the user journey: Rather than shoot the same message to everyone, use a separate sequence for your homepage, product pages, and Facebook Business Page.
- Keep users goal in mind: Emirates bot is more than just a cool ad, it’s the direct line to a desired goal. Users have a task in mind when browsing your site or Facebook Page. Your bot needs to help them complete it as quick as possible.
- Flatter users with frictionless experiences: Imagine being able to book your travel right in this ad, without leaving your current page? Maybe skeptical at first. But considering that travel search is a huge headache, booking “right here, right now” is the most practical and convenient way to do so.
Chatbot Example #2: National Geographic
Ever wonder what it would be like to talk to Albert Einstein? Well, now you can — kinda. To help promote their new show Genius, National Geographic created a conversational chatbot who spoke like Einstein would. While most companies’ blast promos into your chat window, they used Einstein bot to follow users conversation and reply with information about the show and other interesting bits.
The result? 6-8 minute average conversations, 11 turns per conversation, 50% user re-engagement and an involved community of followers. These metrics come from a carefully chosen, user-friendly chatbot strategy.
- Use chatbots to create intimate experiences: People felt they were “talking to Einstein” in real time, in a human and friendly way. Einstein was able to answer professional and personal questions, making users feel a one-to-one connection to the bot, and in turn, the show.
- Make people laugh to engage with them: Laughter is not only good for the soul, but it’s something we don’t expect to do. Especially when you interact with a brand. Build your bot to be informational and conversational, with a little sass here and there.
- Rethink older concepts in a novel way: Chatbots are now used by many different companies across the board. Instead of reusing the same old sequences, try something new that people haven’t seen before.
Chatbot Example #3: Hipmunk
Few niches are more competitive and saturated than travel, but Hipmunk was able to cut through the noise with their location-based approach to booking flights, hotels, cars, and more. Considering the average person visits over 38 sites before booking, Hipmunk is an engaging and convenient way to book travel.
Here’s an example of their Facebook Messenger Bot
The sequence is perfectly tailored to the intricacies of how users search:
- Send deals or promotions based on user preferences: It’s a simple, yet powerful feature that helps users pull information from multiple sites, all in one place — Messenger. Anyone who’s used to the “million tab search method” can finally feel some relief during their search.
- Use conversational language to personalize experiences: Mastering conversation builds trust between bot and user. Each interaction mimics a live agent, making it easy for users to connect with and understand. Imagine talking to a friend or live agent.
- Create content that is rich and useful. In most situations, users wait a bit to get results from different sites. But with ‘HipTips’, their wait time turns into a way to find new deals and ways to use the app.
Hipmunk brings two powerful elements to the table here: proactive outreach and simplifying an exhaustive process for users. You can easily get where you want to go without visiting travel sites and finding deals yourself.
Chatbot Example #4: Lidl’s Winebot
Educated customers are empowered customers, and Lidl’s Winebot Margot brings a warm personality to the stuffy world of wine. Users can get guidance on which wine to buy, tips on food pairing, and even learn about how wine is made. It can also recommend over 220 food pairings and answer questions based on 640 different types of grapes.
- Don’t hide your bot: Margot lives in many different places in Lidl’s eco-system. Most activity comes from their website, which then provides instructions on how to launch the bot in Messenger. Point in case, make it easy it find throughout your company.
- Offer live chat to improve engagement: We know that chatbots can’t solve every problem. Which is why you want to always offer easy access to live support if possible.
- Match the language your audience uses: Margot learns as she goes. The bot has learned how to filter recommendations based on price — but not everyone uses the same lingo. Margot began to recognize the difference between user queries using £5, 5 quid, a fiver, and even “around five-ish”.
Chatbot Example #5: Kia
Despite being part of a complex industry, car maker KIA uses a chatbot that turns Messenger into a viable platform for information. “Kian” has generated 3x more interactions than the corporate site, talks to 115,000 users per week, and has a 21% conversion rate.
These incredible metrics are the result of this Facebook Messenger chatbot suited for user search, and lets Kia personalize service at scale:
- Collect user data to improve retargeting. Instead of letting leads disappear without a trace, your messenger bot should capture customer details and needs. It’ll fuel future sales and re-marketing efforts.
- Offer different services in one place: Capturing lead information is valuable, but it’s not the only thing your bot can do. Kia’s Facebook Messenger bot can also provide roadside assistance, bill payment options, and a store locator to meet live salespeople.
- Replicate human interactions to connect: Kian can answer any type of question for any car model that a live salesperson could.
It’s possible that your chatbot can do well to answer questions as well, but that should be a byproduct of the entire experience. You want to give customers everything they need inside Messenger — from information to payments and more.
Chatbot Example #6: Whole Foods
Whole Foods’ Facebook Messenger Bot is famed for its recipes, product recommendations, and cooking inspiration. Their chatbot helps shoppers find recipes on the spot, and drives traffic to their site from Messenger.
The same principles that work for any marketing channel applies even to chatbots. Identify what resonates with your audiences, and build features to benefit them:
- Make searching for desired information easy: Whether shoppers need to look up a recipe at the store quick, or are browsing from the couch — Whole Foods make it easy for users to find the recipes they are looking for. It narrows down information with filters, gives a preview of dishes, then sends users to their website for the full recipe.
- Talk with users how people talk: One of the biggest challenges in building a chatbot. Whole Foods lets users search for recipes using emojis, just like they’d react to a friends post or laugh at a funny meme.
- Be inclusive to everyone: Personalization is often half-baked and poorly executed, but with careful understanding, it can create an interactive experience. By letting users get diet-specific recipes, it helps promote a culture of inclusivity and wellbeing that welcomes everyone to Whole Foods.
Chatbot Example #7: Madison Reed
Meet Madi, Madison Reed’s genius hair color bot who’s “obsessed with finding your most flattering shade”. This chatbot example is mostly text-based, but what sets it apart from the list of chatbots is the selfie submission. Users can submit a selfie, and Madi will point them towards the best color product recommendations. Cool, right?
Madi teaches us a few important lessons on how to build the best chatbot.
- Speak your audience’s language: Informal tone is one thing, but speaking your audiences style is another. Madi uses “x’s and o’s” as well as flattering comments to build a connection with their audience.
- Solve an actual problem: Similar to most lead generation tactics, your chatbot should help alleviate a struggle your buyers have. For example, waiting for a colorist appointment can take months. Madi is like having a personal colorist available 24-7.
- Ask relevant questions: Madi asks questions such as “What color is your natural hair (not including gray)?” and “Is it chemically treated?” to garner responses from users. She’s witty, funny, and feels like your talking to a friend on Facebook.
Charming always wins. Madi has boosted engagement by 400% with a 21% click-through rate to the company’s website.
Chatbot Example #8: Duolingo
Learning a new language is no easy task. You have the words in your head, but they never come out like you wanted. Turns out — embarrassment when talking with native speakers is real. And Duolingo curved that fear with AI-powered chatbot app.
Although the Duolingo Bot is for language learning, it taught us three lessons for building a chatbot for any industry.
- Build around existing pains to see success: Listen to your users and build sequences based on their pains. Doing this provides a strong incentive for them to use it. Not sure what to write? Simply reach out and talk to your customers about their goals.
- Make it easily available for maximum usage: Shutting down your bot at any time leads to missed opportunities and revenue. Instead, make your bot available at any time of the day, for any reason.
- Design your bot to be interactive and enjoyable: Chatbots that retread the same script, for the same people, don’t make an exciting experience.
Chatbot Example #9: Civilized Caveman
Civilized Caveman was one of the first companies to use a Facebook Messenger Chatbot quiz. Users who took their 1-minute “sugar quiz” were given a 7-day detox as a result. Then the brand could build a relationship with their now subscribers via content and offerings.
Along with conversational tone, the Civilized Caveman quiz provokes an emotion. Whether you are curious or struggle with sugar addiction — this quiz opens the door for different types of people to become subscribers.
- Use a quiz to start your funnel. Quizzes are one of the best ways to engage browsers. Plus, they automatically become a subscriber once they click Get Started.
- Take advantage of images to boost engagement. GIFs and images hold users attention for that extra second. You know, the one you probably need to get that click you want.
Chatbot Example #10: TechCrunch
Want to get more eyes on your content? Of course you do. Take TechCrunch’s Facebook Messenger bot as the main chatbot example for personalized distribution. The bot finds out the type of content you want, when you want it, then delivers on autopilot.
There are a ton of other custom features to go over, or you can check it out yourself to get the jist of it.
- Let subscribers lead the way: Distribution plans only work if timing and context are factored in. Follow TechCrunch off a cliff and let your subscribers choose when, what, and how often they want to hear from you.
- Be open to feedback: Don’t guess what your customers want — just ask. As a way to increase engagement rates and site traffic, TechCrunch lets users suggest new content. So they only read what they want to read.
- Do less, but do it well: It’ll be tempting want to do everything — but start slow. Chatbots who try to do everything at first typically fail.
Since it’s deployment, TechCrunch has continued to make their Facebook Messenger Bot smarter and more conversational. It now understands more natural language queries, conversational double intent, and conversational negation.
Chatbot Example #11: Amtrak
One reason the travel industry has been the fastest to adopt chatbots is speed and convenience. In this example, Amtrak helps site visitors plan vacation, book reservations, navigate Amtrak.com, get route information, and more. It’s become a company powerhouse, generating 30% more revenue per booking, 8X ROI, and answers over 5 million questions a year.
- Leverage post-sales support as revenue channel. Returning customers spend 67% more on average, and more often if they are satisfied with your service. Quick and convenient customer support (in any industry) leads to happier customers, and for Amtrak it’s proved to be insanely profitable.
- Have Live Chat ready for back-up. Chatbots can’t answer everything. Instead of aggravating people, design your bot to get browsers in touch with live support for queries that need a little human love.
- Build your bot around customer problems. For example, more travelers are booking tickets online compared to other methods. Amtrak wanted to make it easy for their 375,000 daily visitors to self-service. Julie was the answer.
Chatbot Example #12: Dominos
Dominos changed the way we order food with Dom, the Interactive Pizza Bot. The Domino’s chatbot can re-order previous orders, offer a full menu, and track deliveries all in Messenger. It’s become one of the best chatbots for customer service — and now I want a pizza.
- Make it accessible across platforms: Chatbots that span cross-platform perform better than one channel. For example, you can order through the Domino’s Pizza Bot by Slack, Messenger, Apple Watch, Mobile App, Twitter (by Pizza Emoji), Smart Home Assistants, and more.
- Don’t be afraid to be fun: More often than not, companies are afraid to be too simple or fun. This will not diminish your brand. Executing a key interaction with an emoji will go a lot further than a wall of text in your chat window.
- Start simple, then move on to more complex cases: What’s the biggest reason users interact with your site or Facebook Page? If it’s ordering a pizza, design your bot around that core function. Once that experience is flawless, build your out chatbot with key secondary interactions.
Inspired by these brands using chatbots? Here’s how to make your own
Ready to make an impact with your own conversational chatbot? If so, you can get started to by doing the following …