Bots…Fad or Frontier — Part 1: The History of Chatbots
By Jason Raziano
Over the past year, the term “chatbot” has taken the media world by storm, sent app developers back to the drawing board, and convinced Fortune 500 companies to entirely reevaluate their customer engagement strategy. But, as with most hot new technology trends that generate a lot of buzz, it’s particularly important to take a step back to recalibrate and break them down into pieces.
So what exactly are chatbots? How come every media outlet is buzzing about them? Are they really such a large opportunity?
Naturally, as a Summer Associate at a VC fund, I decided to take a deep dive into the space and answer these questions. I will walk you through the past, present, and future of bots over a series of blog posts.
This first post will evaluate the history of chatbots and external forces driving this new trend.
It’s worth noting that although there have been several posts specifically covering the history of chatbots, I will add on to what’s out there, highlight the most compelling pieces, and dig deeper into certain areas.
I hope you find this helpful, and as always, feedback, comments, and thoughts are welcome.
Before we jump in, let’s go over the basics.
So what exactly is a chatbot?
Matt Schlicht in Chatbot Magazine defined them rather eloquently: “A chatbot is a service, powered by rules and sometimes artificial intelligence, that you interact with via a chat interface. The service could be any number of things, ranging from functional to fun, and it could live in any major chat product (Facebook Messenger, Slack, Telegram, Text Messages, etc.)”
For those of you 90s babies, you probably remember SmarterChild, the chatbot that was created inside of America Online’s Instant Messenger. SmarterChild, resided in each user’s buddy list, just like any other friend. You could message him for random data inquiries regarding a myriad of topics — stock quotes, movie times, news, weather forecasts, sports, as well as many other topics. As far as I can remember, this was my first engagement with chatbots. SmarterChild was created 16 years ago by a startup called ActiveBuddy, founded by Robert Hoffer, Timothy Kay and Peter Levitan. According to Hoffer, at its heyday, SmarterChild interacted with 30 million people and accounted for 5% of all AIM traffic! It was eventually acquired by Microsoft in 2007, and ultimately faded out of popularity.
So why is the story of SmarterChild relevant?
It’s relevant because the conceptualization of chatbots is not new and specifically, SmarterChild is 15 years old — as is the hype, excitement, and initial craze. Yet, almost a decade and a half later, the term “chatbot” is dominating media headlines and sending VCs scrambling to make investments. What is different this time? Is it market timing? Are there other driving forces? Let’s dig deeper.
The poster child that is frequently cited when predicting how messaging-based/AI bots will transform in the US is WeChat. WeChat is the dominant messaging platform in China owned by Tencent. The massively popular Chinese chat app surpassed 700 million MAU during March 2016, up 29% year-over-year (YoY) from the same period in 2015, according to data from Quest Mobile.
Unlike the fragmented market in the US which is split between the big players such as Facebook Messenger, Snapchat, Slack, WhatsApp, SMS, amongst others, WeChat’s only true competitor is QQ which has roughly 150 million less MAU (monthly active users) and is also owned by Chinese tech titan Tencent. However, what really makes WeChat so fascinating is that it has transformed into such a sophisticated, engaging, and multi-use platform far beyond just messaging.
-Multimedia content messaging/sharing (Pictures, Audio, Videos, Emojis, Stickers, articles, blog posts)
-Voice/Video Calling-Official Accounts (Verified accounts with specific features tailored for celebrities, businesses, etc)
-Ads, Games, Stickers
-In-App Payments (Pay friends, split bills, pay businesses, pay utility bills, healthcare bills, send gift cards, participate in crowdfunding, use coupons, pay a loan, and more)
-Complex search options
A16z’s Connie Chan, put it best: “WeChat: the pioneering model of “apps within an app”. Millions (note, not just thousands) of lightweight apps live inside WeChat, much like webpages live on the internet. There are well over 10 million of these official accounts on the platform — ranging from celebrities, banks, media outlets, and fashion brands to hospitals, drug stores, car manufacturers, internet startups, personal blogs, and more.”
Since WeChat has successfully transformed messaging into the new platform or arguably, the new operating system, the integration of AI, or messaging based bots now comes into play. As Dan Grover puts it: “So let’s take these past few years in China as The Great Conversational UI Experiment.” I tend to agree with him.
So the real question becomes: Is messaging eating software? Will it become the new platform? This is precisely why bots are so interesting now.
Enter: Facebook, Slack, Microsoft, Google, Amazon, WhatsApp, Kik, Telegram…
Since December of 2015, there were several noteworthy announcements from the large tech players — Google, Microsoft, Facebook, and Amazon — outlining the instrumental role that bots will play in their long term strategy.
The timeline below identifies some of the major Chatbot/AI related announcements leading up to 2016:
In addition to the large tech players entering the arena, there are four large external trends powering the bot movement:
AI is Rapidly Advancing
Artificial Intelligence is rapidly advancing and people are now utilizing it to automate daily inefficiencies and mundanities through chat-based interfaces.
Messaging Apps Have Now Surpassed Social Networks
To put this into perspective, between Facebook Messenger and WhatsApp, people send around 60 billion messages every day (almost 3x as many messages as SMS handled at its peak.) Thus, many bots are now being created inside messaging apps (Slack, Facebook, Kik, etc) — and it should feel like you’re chatting back and forth as you would with a human.
Apps are “Dying”
The third, yet more controversial trend is the app problem. Simply put, people aren’t downloading nearly as many apps as before. According to ComScore the average person downloads 0 apps / month vs 50k app submits a month. Additionally, there is a diminishing marginal return for each additional app users download. With the mass adoption of mobile, apps have served as the primary distribution network between consumers and businesses. However, the nature of that relationship is changing — or so it appears. The average global mobile user has = ~33 apps installed on his or her device and 12 apps used daily…80% of the average global mobile user’s time is spent on 3 apps. The friction of getting consumers to download and use an app is becoming extremely difficult and quite costly as cost-per-install and paid acquisition marketing are increasing. I can personally attest to how many “zombie apps” now exist on my phone.
Developers are Frustrated
Developers are at the mercy of the app store and other platforms.
Clearly, there are both micro and macro forces laying the groundwork for chatbots to embed themselves into the daily life of both the consumer and enterprise end–user, and most people would concur. However, the much larger debate stems around the market timing, specific use cases, sophistication, and the overall significance of bots. Will bots have a similar impact to society that mobile apps had? Will voicebots like Amazon’s Alexa outperform chatbots? How advanced are the Natural Language Processors powering chatbots and will they actually work? Kik Co-founder, Ted Livingston, claims “This is the beginning of a new internet”, while others dismiss chatbots entirely — I think we are somewhere in the middle.
The next post in this series will outline the current landscape of chatbots, explain my framework for evaluating chatbot startups, and analyze the specific use cases for chatbots. Stay tuned.