Emojis as Content within Chatbots and NLPs

Emojis are more than just a fun visual graphic 🙌, they convey meaning. And within a chatbot, an emoji can be used by both the bot 🖥💁 and the bot user 👩 to communicate concrete ideas 💡. Let’s look at how emojis can be used by bots and more interestingly IMHO by users.

What is an Emoji?

I am sure you know what an emoji is from the visual sense 😊. But did you know emojis are encoded in the Unicode Standard. Starting in 2010, Unicode has included emojis, which enables them to be widely used.

Emoji is a standardised set of characters that is available on iOS, Android, Windows and OS X. While the artwork for each emoji character varies by platform, the meaning of each symbol remains the same.

-- Emojipedia

What does that mean?

Emojis are equivalent to letters in Unicode, so they are standardized. While they look different on each platform (much the way letters look different in different fonts), they are sent and received by chatbots as text and not as images. This is an important difference (text ≠ 🖼).

Unicode Character Unicode Apple Google Facebook
 A (Latin Capital Letter A)  U+0041 n/a n/a n/a
 Δ (Greek Capital Letter Delta)  U+0394 n/a n/a n/a
 😃 (Grinning Face Emoji)  U+1F603
 👍 (Thumbs Up Sign)  U+1F44D

Emojis Sent from Chatbots

Let's face it, chatbot messages can be boring. If all we see from a chatbot is a series of text messages, it is hard to stay engaged. Along with other visual elements (e.g., Messenger's Button Template and Slack's Interactive Buttons), emojis can help create a more compelling visual experience for the user.

The I will Vote chatbot makes good use of emojis to add some personality to the bot.

From a technical standpoint, adding emojis is simple, since they are just text. Just copy and paste the emoji into your NLP or hard coded responses, and you are done!

Emojis Sent to Chatbots by Users

The tricker and more interesting (to me 😉) part about emoji usage is how to handle an emoji sent from the bot user. Since most of us use emojis (92% says Emogi), we need to start thinking about how a chatbot will understand an emoji.

Let us start with a simple example of trying to teach a NLP (Natural Language Processor) to understand a yes/no response. I will use Facebook Messenger for the chatbot and Api.ai for the NLP.

Setup of Api.ai

First create 2 entities to handle the yes/no answers.

Entity: @Answer-Yes

Entity: @Answer-No

Then we modify the Default Welcome Intent and add two new Intents: Answer-Positive and Answer-Negative.

Intent: Default Welcome Intent

Intent: Answer-Positive

Intent: Answer-Negative

Test in Api.ai

Now that Api.ai is setup, it is time to test within the api.ai console.

Test 1: Positive Response

The first test will confirm that the agent understands a positive/yes answer.

  • In Try it now..., type: hi
  • Response: Is your day going well?
  • In Try it now..., type: yes
  • Expected:
    • Response: I'm glad to hear that.
    • Intent: Answer-Positive
  • note: the Parameter value is Yes

Test 2: Negative Response

The next test is to confirm that the negative/no answer is working correctly.

  • In Try it now..., type: hi
  • Response: Is your day going well?
  • In Try it now..., type: no
  • Expected:
    • Response: I'm sorry to hear that.
    • Intent: Answer-Negative

Test 3: Emoji - Fail

Finally, it is time to confirm that api.ai is unable to understand an emoji. For those into TDD, this is the Red/Fail test

  • In Try it now..., type: hi
  • Response: Is your day going well?
  • In Try it now..., paste: 👍
  • Expected:
    • Response: I'm not sure I follow.
      • or a similar message
    • Intent: Default Fallback Intent

Update Api.ai Entity

The next step is to train the NLP to understand the emoji. This only requires adding a synonym to the @Answer-Yes Entity.


Final Testing

Now that the emoji 👍 is add to the entity, it should work in api.ai and in the chatbot. Onto the final tests.

Api.ai Test 4: Emoji - Success

Repeating Test 3 again in the api.ai console shows that the agent now understands the emoji.

  • In Try it now..., type: hi
  • Response: Is your day going well?
  • In Try it now..., paste: 👍
  • Expected:
    • Response: I'm glad to hear that.
    • Intent: Answer-Positive
  • note: the parameter value is Yes; this is because 👍 is a synonom for Yes in @Answer-Yes Entity

Test 5: Facebook Messenger

Since it all works inside of api.ai, it is time to do the last test within Facebook Messenger.

  • Open the chatbot in Facebook Messenger
  • type: hi
    • response: Is your day going well?
  • type: yes
    • repsponse: I'm glad to hear that.
  • type: hi
    • response: Is your day going well?
  • type: no
    • repsponse: I'm sorry to hear that.
  • type: hi
    • response: Is your day going well?
  • paste or select from emojis: 👍
    • repsponse: I'm glad to hear that.
    • success!

Additional Thoughts

To wrap up, Emojis can be included in the content your chatbot sends to users, and chatbots can be trained to understand emojis.

Easy Part - Use Emojis in Content

If the personality of the chatbot allows it, I suggest using them. Looking at line after line of content can be confusing and boring. Emojis can add some nice visual queues and flare.

Hard Part - Understanding Emojis from User

By there nature, emojis are imprecise. So when asking a question, some emoji answers will be very clear and others will not:

Would you like me to submit your order?

  • Easier to understand:
    • 👍 👌
    • 👎
  • Harder to understand:
    • ⏰🐷✈️ (when pigs fly)

So depending on your audience, it may or may not be a priority. However, from a technology standpoint, chatbots and NLPs are ready.

Beyond the Hype, Chatbots Will Be Serious Business, Part 3: Who Will Help Businesses Succeed with Messaging

By Alex Kaplinsky

Not unlike websites or apps, there is risk for brands in getting chatbots wrong. No one wants to build the next Tay — Microsoft’s teen-friendly chatbot, which Twitter users turned into a crazy racist in record time — nor does anyone want to be responsible for creating Clippy 2.0 and annoying users everywhere. Either building the wrong functionality, wildly misjudging your audience, creating the wrong user-experience, or just having slow crappy performance and functionality (engineering/integration) could make a chatbot project go horribly wrong.

Once you see chatbots as not just a small digital tool companies are adding to their arsenal but as a new communication platform, it’s clear that companies will need more than just an internal developer working on their bots. In much the same way that a website needs more than just one guy on your team who can write HTML, building chatbots are going to require a team and a strategy — and yes, that team will include engineers and designers, too, as well as writers, project managers, and brand strategists.

In fact, I see messaging-focused digital agencies building the same sorts of teams that already exist in digital agencies that work on web and mobile projects today. Here’s how I see the teams of today mapping to messaging teams in the future:

Project Managers as Product Managers

The best digital project managers, whether building a website or an app, carry the skillsets of a product manager. While not necessarily engineers by trade, they must understand business strategy, user-experience, and software engineering. More importantly, they must be able to think of these disciplines in an integrated manner where the end result solves both a business problem and a customer need. As we’ll describe below, the disciplines to make a great messaging app will be varied and thus demand the same integrated capability.

User-Experience as Interaction Design

In messaging apps, user experience will morph to include not just designing visual interfaces but also architecting a text and call-and-response-driven architecture. While much of this will live at the AI level, there will be a large need to structure how you guide and respond to a user through a series of text, commands, and options. User personas will now be matched to “chat personas.”

Calling All Poets and Screenwriters: This Is Your Moment in Tech

Know a poet or screenwriter? Someone who makes great web series, maybe? They’re about to become a hot commodity for the first time in their lives. Brands are often described as having a personality, but nowhere will personality be more important than in chatbots. The goal of a great messaging experience is one that is largely indistinguishable from interaction with a wonderfully helpful and engaging person. Writers with backgrounds in performing arts will help define and create the interaction style and overall persona of every chatbot. Some brands may even opt to have multiple different chatbots, personalized to different types of users. Some will want a funny chatbot, others a super-friendly one, and all will want bots that both align with their brand and help them stand out amid a sea of other bots.


The front-end developer has always had the role of taking the data from the back-end systems and presenting it in an appealing format to the user. The same role will need to be applied with chatbots, but here’s the big shift: instead of using only HTML as the display mechanism and form fields and clicks as the input mechanism, front-end developers will also need to be parsing text for meaning, and then inserting the data from the back end into a meaningful text output to the user.

As chat platforms begin to support more than text — some do already, and all will need to soon — it will be text, images, links, tables, and possibly even full HTML widgets that will be the input and display mechanisms. But the front end developer will still need to take the inputs, pass them on to the back end systems, get back responses and display those to the end user. It will be just as imperative that this is done in strict compliance with the brand voice as it is that on a website the look and feel of the final site matches the visual design handed off to the developer.

Back-end Development Will Still Be Necessary, and Will Still Be about Business Logic and Systems Integration

Back-end developers generally care little about the form factor that the user integrates. Generally, they get inputs, have to search a number of systems and then return results. With the development of AI, machines can possibly do more of this work themselves rather than having to be told exactly what to do, but the first bots will likely have a proscribed set of capabilities that will be codified in the business logic of the back-end systems.

As with today’s modern web- and app-based systems, the back-end developer will be working in virtualized environments where they are integrating cloud platforms that drive content, commerce, loyalty and support systems. The underlying systems and the inputs and outputs will remain fairly constant from today’s frameworks as the same systems will need to support multiple channels with integrated data and business rules.

A Little Less Click Bait

Analytics for chatbots will be less click-driven and more about behavioral responses. Web analytics have become increasingly sophisticated but largely are used to track a user through a traditional funnel. In messaging, analytics teams will continually look at the “middle of funnel” engagement patterns to see where user needs are going unmet and where users are being disappointed. A/B Testing of different personalities, phrasings, and responses will become commonplace. So much so that it will become central to iterative development.

So, while chatbots will be able to avoid some of the development costs of other digital platforms — they’re inherently platform- and device-agnostic for a start — playing in the chatbot space will still require a lot of the same people you need today for web and mobile applications, albeit with some new and different skills.

Want to know more about chatbots? Read Part 1: Why Messaging Is Happening and What It Means and Part 2: How Messaging Transforms Transactions

Beyond the Hype, Chatbots Will Be Serious Business. Part 2: How Messaging Transforms Transactions

By Tim Ross

One way to wrap your head around the chatbot trend is to look at how they’re being used in China, where it’s become a popular commerce tool (people often use chat to order in a restaurant, for example). But as the ecosystem of bots grows in the U.S., I see them handling various other things for businesses, too. Key to their adoption is the convenience for consumers; with chat, you no longer have to:

  • Download a new app
  • Create a new profile
  • Enter payment info every time you want to purchase something on mobile.

Instead, your profile, payment info, and location will be stored in your chat app, enabling quick and easy transactions with a variety of chatbots.

Specifically, I think we’ll be seeing chatbots rule the following transactions:

Content & Information

Customized Content

This is an obvious one, but worth taking a closer look at nonetheless. There are a variety of ways in which chatbot content will be a useful customer acquisition tool. First, bots will serve as a more effective, natural language version of search. Users will be able to ask questions and receive automated answers and information. This can range from product info and comparison queries, to suggested locations, to instructions and tutorials, and more.

Then, content can be customized based either on information a business already knows about a customer (e.g. where they live), or on answers to a few questions.

To ensure a positive, personalized experience, a lot of chatbots will start users off with a survey, as Sephora’s Kik chatbot does.

Just think about the myriad interactions this will change. If you get emails about new real estate listings right now, for example, imagine getting chats about them instead, and being able to ask questions about particular listings or request to see more of a particular type of listing. If you’re traveling, you might get instant, guidebook-style info on where to eat, stay or shop. Pharmaceutical ads? Way better in chat, where you can ask questions and seek information privately, and read that long list of side effects to yourself. Want to explain to a customer how one of your products compares to another, or how your product compares to your competitor’s product? It’s much easier to do in a way that will actually get the message across in chat than it is in an ad.

Speaking of ads, chatbots will likely do away with the need for those terrible, long custom URLs pegged to every TV or print ad. Instead of expecting people to remember these or write them down, advertisers can switch to a simple chat address. Again in the context of a pharmaceutical ad, someone might see an ad for a particular drug, easily engage with the company on their phone and use chat to ask questions and get a diagnosis of sorts. The chatbot could serve up content that’s customized based on the user’s questions, and guide them toward various calls to action.

Chatbots can not only be programmed to have a specific “personality” but also to respond to users’ behavior in ways that make sense — noticing when a user is getting annoyed, for example, or when they seem to be excited about a particular product or service and may want to learn more.


If you come up with a chatbot that does something cool or provides a unique experience above and beyond the quotidian interactions between any customer and company, you’ve created a powerful new customer acquisition tool. People who may either be unfamiliar with your brand or even dislike your brand could feasibly interact with your bot and change their minds. Moreover, early research is indicating that chatbots are far better than any other digital tool at actually getting online customers into offline stores and at converting digital interest into actual dollars. Chatbots in Kik and Snapchat, for example, offer specific codes that customers can use to order at restaurants or redeem coupons.

Advertising and sponsored content are also making their way into chat apps. Although Facebook is so far not offering that option, Snapchat and Kik both include video ads, and Snapchat allows brands to sponsor channels in its Discover feature. Kik even allows businesses to start conversations with customers instead of waiting for customers to contact them (it remains to be seen whether customers will decide this is just annoying). The one thing that may save ads from suffering the same fate in chat that they’ve suffered in other media — namely that people find a way around them — is the ability of chatbots to contextualize content; so an add that shows up in a particular chat stream will actually make sense and have a higher chance of achieving its intended purpose.

Activating Influencers

In the same way that brands have linked up with Instagram sensations and Twitter kings, they will link up with key influencers to integrate chat promotions, or even to create their own chatbot persona for users to interact with. This is already happening in WeChat in China, which means it’s only a matter of time before we start to see it stateside.


In many use cases, messaging will represent a faster, more frictionless commerce vehicle. But chat-assisted commerce will be most effective with relatively simple purchases in physical environments where the reduced friction increases purchasing volume — a restaurant, for example, where it might be easier to sit at a table, chat your order and have it magically appear than wait for a server and go through the whole ordering process (see the example to the left, from a restaurant in China with a WeChat chatbot). I don’t think it will be long before chatbots are being used to skip lines in crowded businesses, either, or to place to-go orders, room service orders, or purchase specialty items (pastries at your favorite bakery, for example) ahead of time. Enabling these types of solutions will require integration into CRM, transaction, fulfillment, routing and more.

In-Store Support & Assistance

Once you’ve got your customer into your off-line store, you can also use chatbots to deliver a customized physical shopping experience, notifying them of deals on their favorite items, and helping them to find them in a store. As more and more chatbots are integrated with POS systems, customers will also be able to pay and check out, using a chatbot instead of a cashier and skipping the line altogether.


Stores have been investing in indoor navigation and wayfinding for a few years now, leveraging beacons and mapping information to help customers find their way through stores and to suggest products along the way. That functionality has never been exactly seamless, though, and has always required the downloading of yet another app. Chatbots could improve on this experience, helping customers navigate, offering information, and answering questions throughout a customer’s retail experience.

Imagine a shopper in Target, for example, who could ask about where to find a particular product, and be served up not only directions to that aisle in the store, but also information about that product and other related products, or even pushed a coupon for the product in question.

Travel Reservations

There’s not a single airline, hotel, or rental car app out there that customers enjoy using. Instead, chatbots from these companies could easily interact with customers, enabling them to make, change, and cancel reservations, or request additional services or upgrades.

Venue Information & Commerce

Similarly, a great mobile experience does not yet exist for booking or managing concert or event tickets, but chatbots could go a long way toward addressing that gap. Today the experience goes roughly like this: download app, create an account, enter payment info. In the chat-based version of this interaction, it would go more like: open chat window that’s already on your phone, order, pay with existing credentials. Your profile and payment info are already stored, and the chatbot will also generally know your location, which means it can recommend events and better support your purchase.

Customers could find information about and directions to venues through chat and promoters can keep them posted on any changes to the event they’re attending, or deliver important information about what to wear or bring, when to arrive, and so forth. Those who opt in to push notifications could get messages about events happening nearby when they’re out and about, or about a concert coming up with a band they like or have seen before. When they get to the venue, customers could use chat to figure out where the nearest concession is, or even order ahead so they don’t have to leave their seat for too long. Just imagine having a beer brought to your seat during a ball game, or not having to wait in line for food at a concert.

Gamification & Entertainment

Games and quirky, fun experiences are a natural addition to interactions in chat, in a way that they just couldn’t be in real life. You probably wouldn’t interact with someone dressed up as Miss Piggy in Union Square, for example, but having a chat with Miss Piggy? It’s kind of fun. And it really helped Disney promote its Muppet Show on ABC.

Loyalty & Offers

For marketers, messaging represents a huge opportunity for promotions and loyalty marketing. Loyalty is something people have been trying to figure out on mobile for years and it’s never quite gotten there — consumers don’t really want to download a different app for every store’s loyalty program, and aggregator apps have failed to get enough businesses on board to make consumers feel like they’re really worth it. By integrating with CRM and digital direct-response systems, messaging can provide an instant, context- and location-aware vehicle to push personalized offers to users. But unlike the used car salesman feel of coupon offers, natural language bots can make this feel much more like a concierge at a hotel, or the greeter at a favorite restaurant.

Meanwhile, targeting promotions to a company’s chat followers can be useful in figuring out what sorts of consumers are engaging with you on chat, and what they’re likely to respond to.

Offline Offers

Connecting digital customers to offline offers or actions has always been a challenge. I’m not going to go so far as to say that chat solves that entirely, but it goes a long way toward closing that gap. A customer that has opted in to push notifications, for example, could get a “welcome back” message the next time they’re in a physical store, with information about what’s on sale or what offers they’re eligible for.

Appointment Openings & Upgrades

Wellness, fitness, and beauty businesses have lagged behind other industries on the digital front, but chat could go a long way toward improving their booking capabilities for customers and their inventory management for business owners. Say a customer uses chat to book a massage at a spa, for example. On the day of their appointment, if the spa had additional appointments available, it could message the customer asking if they might be interested in adding on a facial or pedicure. Similarly, say a yoga studio has openings in its evening class — its chatbot could notify regular customers and push notifications to any customers in the vicinity of the studio.

Customer Support

Everyone was excited when support got more social, but messaging is another big step forward in improving support while reducing costs. By instantly answering questions, a bot can deflect a costly support call and speed time to resolution. Moreover, the bot can provide a more “human” style of interaction and use language processing to assess a customer’s frustration and understand when to escalate cases. By integrating with knowledge bases, CRM and trouble ticket systems, bots will also be able to improve service for ongoing cases, providing continuity and keeping consumers in the loop with status updates.

Trouble Ticket & Follow-up

Concierge support services like Zendesk and CRM offerings like Salesforce are already looking at integrating with bots. Facebook Messenger chatbot launch partner Spring integrated with Zendesk to provide a hybrid chatbot-human experience to shoppers using its Messenger experience. The Spring Messenger chatbot can suggest items to shoppers within Messenger and, if a customer has a specific question — how do sizes run in this brand?, for example — automatically connect to Spring’s Zendesk live concierge team to answer the question, then seamlessly hand the transaction back to the bot to close the deal.

Early Diagnosis

One of the most valuable uses of chatbots in the Customer Service realm will be their ability to diagnose and address problems before a customer would have traditionally called or emailed the support desk. This will deflect calls, keep customers happy, and reduce support costs at the same time.

Appointment Scheduling

Whether it’s a repair appointment, a consultation, or a service appointment — a haircut, for example — scheduling is much easier through chat than through email or phone. Rather than call, email, or have to use one of the many broken online scheduling widgets, customers can simply chat an appointment request, respond to the bot with the date and time that works for them, and that’s that.

Reminders & Follow-ups

Chat is also a less intrusive way to remind people that they have an appointment coming up, and a more convenient way to collect post-visit feedback. Just think about every time you go to a store or deal with a customer service rep and are asked to either go online and fill out a survey or stay on the call to take a survey. A very small percentage of people actually do this, and we all know that they are not representative of the general public. A quick chat survey, though? Many more customers are likely to engage with that.

Reservation Confirmations

Imagine if you got a reservation confirmation via chat instead of a phone call, or an email that goes to your spam folder. That chat could include simple “reschedule” and “cancel” buttons to make it easy to change plans, and the restaurant’s chatbot could even tell you about that night’s specials before you get to your table.

Travel Logistics

Some hotel chains (Hyatt, for one) and airlines (KLM) are already using chatbots to perform a variety of functions that have always been clunky on other digital platforms: reservation confirmations, check-ins, and special requests. Instead of downloading the mobile app for an airline or hotel — which means eating up data for what generally amounts to a frustrating experience — travelers can keep tabs on both hotel and airline reservations from the same chat app, and use various call-to-action buttons to check in or make changes. The number of hotels and airlines going this route, and the variety of services they offer via chat is only going to increase in the coming months. Requesting a seat assignment on KLM long before you’re standing at the ticket counter, for example, or checking in to your room at the Hyatt in the morning and letting them know you’ll be arriving late … maybe even ordering room service as you’re in a cab on your way.

Ultimately, for businesses bots are an excellent way to engage customers. I talk about chatbots in the context of conversational engagement because I think that’s their biggest value to companies — not only providing the services mentioned above, but doing so in a way that continually improves a company’s relationship with its customers.

Want to know more? Continue reading Part 3: Who Will Help Businesses Succeed with Messaging

Beyond the Hype, Chatbots Will Be Serious Business, Part 1: Why Messaging Is Happening and What It Means

Facebook’s Messenger announcement at its annual developer conference shows a depth of seriousness in platform, focus and ambition around messaging as a legitimate communication channel for businesses. But where did the chatbot revolution come from and why is it picking up steam now? It’s simple really: As more and more consumers are transitioning over to mobile for the bulk of their Internet usage, there’s a new digital shift happening from apps to chatbots. Or as Wired puts it: bots are going to power a “post-app Internet.” So what are chatbots exactly? They’re basically software designed to automate tasks and communication. Thanks to advances in artificial intelligence and natural language processing, they’ve advanced beyond the command-line interface they were in the 90s to provide an experience similar to what you’d get if you called a 1–800 number for customer service. Another way to think of it is that chat apps (also known as messenger apps) are essentially the new web browsers, and chatbots are the new websites.

In his keynote at Facebook’s F8 developer conference, CEO Mark Zuckerberg said Messenger is the company’s fastest growing app — faster than Facebook itself. In the last 12 months, messages between business and people in Messenger have doubled. Companies like Walmart, Sprint and Hyatt are using chat for customer service; KLM was an early leader, using chat for ticket confirmations, check-in, and everything in between; and Uber and Lyft have enabled users to request a ride in Messenger without downloading an app. Many people have talked about chatbots replacing 1–800 numbers, and Zuckerberg referenced that, too, noting that 1–800-Flowers has been an early adopter of chat as a commerce tool. “In the near future, to order from 1–800-Flowers, you’ll never have to dial 1–800-Flowers,” Zuckerberg said.

The “not having to download an app” part is the real value of chatbots for consumers. Rather than having to download different apps for every brand they interact with, consumers can download one messaging app and then interact with various company chatbots within that app.

In a special session on Messenger at F8, Facebook’s Jessica Lee described the chatbot revolution as a huge paradigm shift. “Businesses are creating new, conversational experiences,” she said.

To help them do that, Facebook has released a chatbot API and various tools, enabling developers to build chatbots for Messenger. The send-receive API is the backbone of Facebook’s new API tool. It can be used to create automated responses, and also to loop in actual humans when necessary. Message templates allow for chats that can be text, images, or a bit of both, with a call-to-action button.

Lee cited the shopping app Spring as an example, noting that the general mobile shopping experience today is tedious. Spring, a launch partner of Facebook’s Messenger chatbot program, used the API and conversational interface to improve the experience. Other Facebook Messenger chatbot launch partners include Poncho the weather cat, which is using natural language processing to gauge users’ attitudes and respond in kind, and CNN, which is delivering a quick summary of the day’s news to users via Messenger. And in the same way that people can follow or like a business from a variety of places, they’ll be able to engage a company’s messenger bot from that company’s site or from within Messenger.

The point is that chatbots today are not the chatbots of a few decades ago, which were clunky command-line text interfaces that came nowhere near approximating human dialogue. Chat is nothing short of the next digital revolution, providing companies with a unique opportunity to engage consumers conversationally, building better customer experiences and deeper loyalty than ever before.

Why Now?

Facebook is not the only, or even the first, entrant to the chatbot market. Slack already has support for bots, most notably the Taco Bell bot, which enables hungry tech workers to order lunch from within the messaging framework. A week before Facebook’s announcement, teen-favorite messaging app Kik launched its bot store, including bots from H&M and Sephora. And Microsoft has been big on bots for a while as well. While its embarrassing, racist bot debacle was a good example of there still being work to do on the artificial intelligence front, the company’s Chinese bot, X.ai , has none of those issues and has been a popular way for businessmen in China to handle their schedules for months now. In fact, China has been out ahead on this stuff for years, using bots in WeChat to do everything from shop online to order at restaurants. More than 90% of mobile users in China use chat in this way.

Chatbots have become a big thing now for a few reasons. First, people are already using chat all the time. More than 900 million Facebook users already use Messenger, for example, and between Messenger and WhatsApp alone, people are already sending 60 billion chat messages a day. Gartner predicts that by 2020, customers will manage 85% of their relationship with any enterprise without interacting with a human.

Then there’s the fact that the various technologies necessary to make bots any good — primarily artificial intelligence and natural language processing — have matured to a point where they are market-ready. Sure, if you are looking for an AI bot that you can ask open-ended questions and it can automatically figure out what you want (think Siri), then we’re still a few years away. In fact, even in her F8 presentation, Facebook’s Jessica Lee admitted that it’s still early days for bots and that many companies will initially opt for some sort of hybrid between humans and bots. Still, if you are looking for an AI bot that you can ask open-ended questions and it can automatically figure out what you want (think Siri), then we’re still a few years away. But if you’re using natural language processing to create questions that are applied to specific answers (pre-written or at least templated) in a narrow domain? That process works really well today, and it’s a game-changer for businesses hoping to engage mobile customers.

It’s not just that chatbots offer a more convenient, app-free experience for consumers, but also that they actually improve consumers’ interactions with companies. They’re a communication platform, providing what I’ve taken to calling “conversational engagement.”

Every time a consumer logs into their messaging app, for example, they might get push notifications they’ve opted in for — e.g. a Target chatbot might tell a customer that’s in a physical store,

    “Welcome back to Target, Tim, We have a 50% discount sale today on diapers. We know you can never have enough.”

And if they want to engage with a particular company, it will be as easy as selecting that brand’s chatbot from a menu.

The adoption of smart watches, wearables, and beacons, and the increased popularity of voice messaging as well (especially as that technology continues to improve) will only further accelerate the rise of the chatbot. Small (or non-existent) screens demand voice and simplified, new UIs, and beacons enable proximity-based interaction.

Are All Bots Created Equal?

Here’s where it gets interesting: If consumers are using a single app filled with dozens of chatbots, brands are going to need to differentiate their bots. The nature of chat is that it’s casual and conversational, which makes it easier to engage customers without annoying them. But it would be weird if, say, the Sephora chatbot sounded exactly like the CNN chatbot.

Ultimately, I see chatbots as an extension of a company’s brand. They’ll need to have a distinct personality, and that personality may even change a bit depending on what sort of customer they’re talking to. Brands should be seeing chatbots not as yet another digital trend they need to hop on, but as a useful extension of their brand that enables regular, real-time, conversational engagement with their customers.

Chatbots also represent that rarest of things in digital marketing: a communication platform with pretty attractive development economics. Because chatbots are inherently cross-platform and cross-device, without any sort of version or release issues, they’re pretty straightforward to develop and maintain.

As with other digital tools, the challenge here will be both technological (the need for continued improvements in both AI and NLP) and creative. Just as websites and apps started out a little hokey and clunky and evolved into fairly sophisticated experiences, chatbots will evolve to be not only an appropriate extension of a brand, but also to provide experiences that engage, or even delight, customers, and stand out from a herd of other bots.

Which brings us to Part 2: The business use cases.