The Comprehensive Guide to Chatbots

Here is our comprehensive guide for every marketer — and every person — who ever wanted to know more about chatbots.

Chatkit
The Conversationalist by Chatkit
9 min readApr 3, 2018

--

You’ve undoubtedly heard the word “chatbot” being thrown around more and more over the last couple of years. Perhaps you have some understanding of what a chatbot is — you might know it has to do with AI, or machine learning, or you might simply know it’s a way for brands to engage directly with millions of people without their physical presence. Or maybe you have no idea what a chatbot is at all.

Either way, you’re probably wondering: what exactly is a chatbot? How does it work? How does it fit into your larger marketing strategy? And how can it help you cultivate long-term customer loyalty and positively impact your bottom line?

We’re going to address all these questions and more in this comprehensive guide to chatbots. By the end of this guide, we aim to provide you with a clear understanding of what chatbots are, what they do, how you can build and deploy one, and what a massive opportunity they are for your business.

What is a Chatbot?

The Oxford English Dictionary defines a chatbot as a computer program designed to simulate conversation with human users, especially over the Internet. In other words, chatbots mimic human conversation using artificial intelligence, or AI.

AI robot chat might sound like every terrifying sci-fi cliché of humans vs. the machines, but chatbots have actually been around since the 1960s. The first chatbot was named ELIZA, and was developed by MIT professor Joseph Weizenbaum in 1966. ELIZA even passed Alan Turing’s famous Turing Test, which determines whether or not a computer program can impersonate human intelligence.

But let us make it simpler for you.

The best way to think of a chatbot is as a virtual assistant that lives on messaging channels like Facebook Messenger. Using the power of AI, chatbots enable brands to engage with with customers, gain their loyalty, automate routine tasks, provide customer service, and more.

What Chatbots Look Like

Since chatbots live on messaging channels, conversations with bots look like any other conversation you’d have over a messaging app. They’re direct and two-way, and they’re personalized by the first-party data that only a 1:1 conversation can deliver. We’ll explore data more in a bit.

The beauty of dealing directly with customers is that you, the brand, appear in a customer’s buddy list, alongside their real-life friends and family. Chatbot conversations are also ongoing, as is the nature of messaging platforms, and they enable customer and brand to message each other at any time. That means you can message your customers at any point in their purchasing cycle, whether they’re considering a purchase, they’ve left an abandoned cart on your website, they’ve recently made a purchase, or you’re upselling them post purchase. And they can message you back anytime, too.

There’s huge power in that.

This is what a chatbot conversation looks like:

Some Examples of Chatbots

The Danette May, fitness influencer Danette May’s daily motivation bot.

Meditatebot, a personal assistant that guides users through meditation exercises and schedules daily meditation reminders for them.

SnapTravel, a hotel-deal bot that delivers recommendations and enables you to book hotels right on Messenger.

Duolingo’s virtual language-learning tutor, which helps users learn a new language with a bot. Users can even choose the persona they interact with.

Poncho, a weather bot that sends fun and personalized daily weather forecasts over messenger.

ShopBot, eBay’s personal-shopping bot.

Instalocate, a flight-tracking app that not only tracks your flights, but also helps travelers receive compensation for delayed flights.

Why Chatbots

Chatbots solve a lot of the contemporary marketing issues that other digital tools — like email or social media or cookie retargeting — address, but don’t quite crack.

The Internet has removed a lot of the barriers to entry for new products and new companies, and the globalized economy has made it easier for products to compete across different markets. In many ways, this is the best time in history for brands and businesses. But the Internet has also made brands more elusive than ever. Social newsfeed and email inboxes are oversaturated with thousands of messages trying to grab people’s attention, and both customers and email platforms have become more adept at segmenting personal messages from branded ones. Consequently, it’s more difficult than it’s ever been to capture people’s attention.

Email has the advantage of being a direct channel into people’s inboxes, but it’s most often headed straight to junk. Even when people do open emails, they have no way to respond, and brands have few opportunities to capture crucial customer data. Given how quickly people make purchasing decisions, and how immediately they expect their questions to be answered, static channels like email just don’t cut it.

Social media is more of a two-way channel than email, but it requires a lot of human presence, and it’s not intimate or direct. Furthermore, Facebook, the most ubiquitous social network, has pushed branded Pages to the bottom of the Newsfeed in order to privilege interactions between offline friends and family.

Like social media and email, cookie retargeting yields low conversion rates and is difficult to measure, if it even pays off at all: according to the advertising “Law of Seven,” people need to view an ad seven times for it to make any sort of impression. That requires trailing people around the Internet, page to page, likely irritating them in the process, with the dim hope that they click through to your website or landing page.

This is why chatbots have the advantage over every other digital marketing channel. They’re direct, like email, but they boast 90% open rates — 3x better than email. They’re conversational, like social media interactions, but that conversation is 1:1 and endless, and it is personalized based on first-party data, data that comes directly from the customers themselves. And chatbots, unlike retargeting ads, are “forever ads” that keep your brand top of mind the entire funnel, because you’re able to continually re-engage customers.

There’s no need to send them to a landing page or through a pre-planned journey on a website. The entire marketing and purchasing journey can be processed through a chatbot. Your bot can answer common customer questions, patch in a customer service or sales rep for more complex assistance, recommend items, process credit card payments, send you shipping notifications, and follow up after purchase. Chatbots have fundamentally altered the way brands engage customers, sell to them, and retain their loyalty over the long run.

How Chatbots Work

It’s important to deploy a chatbot as part of a larger Chat Marketing strategy, one that is connected to your overall marketing, sales, and business goals. Whether you’re looking to generate leads, increase brand awareness, cultivate customer loyalty, or drive sales, a Chat Marketing strategy ensures your bot is working towards specific goals and not just starting conversations for the sake of conversation.

With a chatbot, you can message your audience at any time with free direct messages or paid sponsorship messages. That means you can send people messages they actually read — messages they want to receive. When a customer clicks on an ad, a parametric code, a link, a widget, a Facebook comment, or any other avenue into a chatbot, they’re automatically guided into a direct conversation with your brand.

Here are some examples of what those conversation starters might look like:

A conversation prompted by a Facebook ad
A conversation prompted by a Facebook comment
A conversation prompted by a parametric code
A conversation prompted by clicking the “Send to Messenger” button

Once you’ve initiated the conversation, there are three steps to connecting through a chatbot:

Step 1: Brand messages customer.

When a customer opts into your chat, you can send them interactive conversations and personalized messages at any time.

Step 2: Customer receives notification.

A notification appears on a customer’s mobile device, like an intimate and personal message from a friend.

Step 3: Customer engages in conversation.

Your brand appears on your customers’ buddy list, which gives them 24/7 direct access to you.

Designing Chatbot Conversations

Once you’ve guided your customer into a chatbot conversation, the possibilities for messaging people are limitless. Here are some best practices for designing chatbot conversations that will represent your brand, engage customers, keep them engaged, and respect both Facebook (or any other chat platform) and your customers:

  • Be careful not to spam people with too many messages. No one wants to be annoyed by an overzealous bot. Instead, your messages should obey the 24+1 rule, which states that bots can only respond to users who initiate conversations. Bots have a 24-hour window to respond to a prompted conversation, and can respond once more outside that window for reengagement.
  • Be upfront with your audience that they’re talking to a bot, not a human. Provide people with clear instructions on how to connect with a human agent, whether it’s through a button, a key phrase, or another prompt.
  • Make sure your chat interface is representative of your brand voice. Use gifs, emojis, and Internet or messaging slang where it makes sense for your brand. If you already share a lot of visual content, use images and other rich content to spice up the conversation. Here is an example of how to imbue your chatbot with your brand’s personality.
  • Don’t try to fit all your content into your welcome bot. Identify the topics you think will be most pertinent and useful for your users and stick with that. Too many options is certain to overwhelm people.
  • Use graphics that are correctly sized for the platform. Facebook Messenger, for example, uses a specific image ratio for images in carousels and card components (1.91:1).
  • Large walls of text can be overwhelming. Break up longer messages into shorter chunks with images, videos, and interactive questions.

Chatbot Use Cases

Here are some of the best applicable uses for chatbots:

Here is an example of what a chatbot for e-commerce can look like.

Chatbots and Analytics

As we’ve mentioned, one of the biggest advantages of chatbots, their differentiator from virtually any other marketing or engagement channel, is that they provide you with first-party data you can track, segment, and understand — and, most importantly, data that you — the brand — owns. Chatbot providers don’t sell your own data back to you: instead, they help you parse your data and use it to personalize customer interactions.

That enables you to know exactly who your data’s coming from and what those people want. For example, if you’re an activewear company, you know people’s sizes, what colors they prefer, what activities they engage in, and where they live. That means you can message them about new summer styles in spring, or winter styles in fall, and you can inform them of any brick-and-mortar popups near them. It also means you know what hours they’re likely to respond. Thus, your data is working to make your bot smarter, more intuitive, and better able to deliver a top-notch customer experience.

Conclusion

Chatbots are fundamentally changing the way people interact with brands. They’re helping brands learn which messages resonate with people at what times, and they’re enabling brands to engage in deeper, more significant conversations with their customers. Chatbots aren’t just a marketing tool: they’re a holistic, end-to-end customer experience channel that can help revolutionize a brand’s sales, customer service, R&D efforts, and more.

Learn more about how Chat Marketing can help your brand at Chatkit.com.

Subscribe to our newsletter and receive more great articles like this one!

--

--