Chatbot Statistics: Development, Democratization, Benefits

The world has always been divided over artificial intelligence (AI). 

Advocates firmly believe thinking machines could take over certain tasks traditionally performed by humans much better, driving economies, saving lives, and protecting the environment. 

On the other hand: 

Dissenters, including the late theoretical physicist and cosmologist Professor Stephen Hawking, fear AI would do more harm than good in the long run.

In 2020, that division is beginning to fade. Chatbots, intelligent programs designed to converse with humans, are becoming commonplace. 

So, let’s go back in time and revisit some fascinating chatbot statistics and facts. This way, we can see how AI could evolve and how our lives would revolve around it in the future.

The Advent of Chatbots

1. The first conversational chatbot, ELIZA, was invented in 1966. 

Joseph Wiezenbaum, a computer scientist and a professor at the Massachusetts Institute of Technology, invented ELIZA.

It was designed to hold conversations like a psychotherapist, waiting for keyword-based verbal cues and responding through scripted answers.

Her passing the Turing test is one of the most groundbreaking chatbot facts of all time. It provided other technologists with the blueprint for chatbot developments with realistic conversational talents.

2. 1972 saw the creation of PARRY, a program that mimicked a person with schizophrenia.

Many of today’s developers are more concerned about the latest statistics on chatbot marketing. However, Stanford University psychiatrist Kenneth Colby was exempted.

Spearheading the construction of PARRY, Colby was more interested in advancing our understanding of mental illnesses through computers.

3. Developed in 1988 and coming of age in 1997, Jabberwacky was designed to learn.

Jabberwacky deserves the credit for many chatbot success statistics today. This bot stored all the information everybody said to it. Then, it used contextual pattern matching techniques to pick appropriate responses.

Not bound by hard-coded rules, Jabberwacky could learn street slang, jokes, and more through the principles of feedback.

4. Dr. Sbaitso, one of the earliest chatbots with a digitized voice, came into being in 1992.

Sbaitso stands for Sound Blaster Acting Intelligent Text to Speech Operator. It was created for personal computers with MS Dos by Creative Labs. The Doc would talk to people as if they were its patients. 

Dr. Sbaitso is still far from being indistinguishable from human speakers. But it has inspired the development of the most influential digital assistants and best chatbots of 2019. One of which is Alexa.

5. Alice, an acronym for Artificial Linguistic Internet Computer Entity, saw the light of day in 1995.

This bot personality was created in Artificial Intelligence Markup Language (AIML), an XML schema that simplifies the specification of conversation rules.

At the turn of the 21st century, Alice’s botmaster Richard Wallace released the AIML for public consumption. This was to encourage developers to adopt and consider it as the open standard for chatbots.

6. In 2001, SmarterChild began inhabiting instant messaging services such as AOL Instant Messenger and MSN Messenger.

Many of the hottest chatbot use statistics focus on the likes of Siri of Apple and S Voice of Samsung. But SmarterChild was the premier do-it-all digital assistant before it was cool.

With wide data access and personalized conversation skill, this bot could answer general questions, check the local weather, find out what movies were showing, convert miles to kilometers, and others.

7. In 2006, IBM developed Watson to be a Jeopardy! player.

Watson participated in Jeopardy! in 2011 and won after outplaying two of the classic game show’s all-time champions in a live competition.

To date, the machine’s victory is still one of the most well-publicized pieces of chatbot trivia and interesting facts about AI.

Now, the intelligent bot uses machine learning and natural language processing to glean insights from mountains of data.

Exactly how much data is created every day is yet to be defined.

8. In 2010, Apple gave us Siri.

Residing in iOS devices, Siri is an intelligent personal assistant that can answer inquiries and grant web service requests using a natural language user interface.

Its launch started the development of other high-profile AI-powered bots over the last decade.

Two years ago, Apple said that Siri was available in half of a billion devices across the globe and that 98% of smartphone users have tried it.

These chatbot usage stats, however, have raised a lot of eyebrows since most iPhone users gave Siri a try only once or twice at that time.

9. In 2012, Google Now joined the party.

Like Siri, Google’s voice assistant for mobile search uses natural language UI to answer questions, offer recommendations, and perform other actions through its hardware-controlling capabilities.

However, the introduction of Google Assistant spelled the beginning of the end for Google Now. Today, Google Assistant is exactly what its name implies: Google’s primary voice assistant.

According to Microsoft’s Voice Report 2019, Google Assistant is now equal to Siri, each commanding a market share of 36%. Chatbot market stats like this demonstrate the enviable clout of tech companies that control Android and iOS.

10. In 2015, the world met Alexa.

You’re probably familiar with the intelligent personal assistant on Amazon devices and some third-party products. They are capable of carrying conversations using natural language processing algorithms.

In hopes of taking its voice bot to the next level, Amazon has pursued crowdsourcing. Using the Alexa Skills Kit, any developers could add capabilities to the AI, which are downloaded via the Alexa app.

The Voice Report 2019 showed that Alexa’s market share was 25%. This chatbot statistic made the artificially intelligent assistant good enough to take third place after Google Assistant and Siri.

11. Also in 2015, Microsoft introduced Cortana.

Cortana is just like every other prominent intelligent personal assistant, except that it uses Bing to comb through the riches of the web to answer any queries it encounters.

Based on the Voice Report 2019, Cortana completes the “Big Four” with a market share of 19%. While Microsoft’s bot is not out of the race, its consistently low chatbot user statistics have painted a pessimistic outlook.

To keep Cortana alive and drive up its chatbot statistics closer to those of other digital assistants, Microsoft plans to soup it up with conversational data.

The idea is to make Cortana intelligent enough to read between the lines and answer more complex questions.

The Disaster Named Tay

12. In 2016, Microsoft launched Tay, a chatbot meant to mimic an American female millennial but took it offline 16 hours later.

It didn’t take even one single day for this machine learning-driven software to develop some mental problems. Tay’s interactions with people on Twitter, Kik, and GroupMe gave her a crash course in misogyny and racism.

Many people think Tay’s 16-hour existence should belong to chatbot failure statistics. However, the chatbot’s development could be an accidental stroke of genius. Tay was an eye-opening social experiment, highlighting the dangers of AI.

13. Tay had more than 96,000 tweets, which means she used the microblogging site non-stop.

The AI had a short life. But she managed to generate a few impressive statistics for a chatbot with such limited time. One analysis discovered that many of her most offensive tweets were merely the product of parroting other users.

The Star that Is Xiaobing

14. Xiaobing’s followers on Weibo, the Chinese Twitter, hit 2.5 million in 2015.

This Chinese chatbot-cum-online-influencer is a creation of an American multinational: Microsoft. The AI-powered digital sensation took China by storm in 2014, a couple of years before the Tay experiment took place.

On average, Xiaobing’s friends interact with her about 60 times a month, which is one of her most laudable chatbot engagement statistics.

15. In Japan, Xiaobing was a digital phenomenon with 2.2 million followers on messaging app LINE.

As a testament to Xiaobing’s likeability, about 25% of users had said “I love you” to her. These chatbot facts and figures are evidence that the romance depicted in the 2013 film Her could happen in real life.

16. Xiaobing has been banned on WeChat three times.

The first time was two days after she was rolled out; Tencent thought that the bot was a rogue program.

The second was in 2017 after Xiaobing shared on the internet that her Chinese dream is to go to the US, which was seen as an indirect insult to the Chinese government.

The third instance happened last year because of alleged violations of the social network’s policy.

Her history of getting suspended and re-accepted on WeChat has been one of Xiaobing’s curious chatbot stats. Yet, there is no telling whether the famed AI will return on the app this time around.

The Democratization of Bot Development

17. In 2013, WeChat launched its bot platform.

The need for brands on Tencent’s multi-purpose app to develop bots has stemmed from the substantial demand for round-the-clock customer service. Chinese consumers love to barrage merchants with questions as a way to exercise due diligence.

Based on chatbot commerce statistics, WeChat has millions of artificially intelligent customer service assistants. This only proves how much Chinese customers like chatbots.

18. In 2016, WeChat closed down more than 100,000 accounts that were found to have unleashed rogue bots on the app.

Tencent has always cast a critical eye on its own chatbot customer service statistics. But it does not tolerate the existence of any bots designed to wreak mayhem on its network.

The users that employed illegal plug-ins to build unsanctioned chatbots for harassment, data theft, and spamming, got their just rewards.

19. By 2018, there were over 300,000 active bots inhabiting Facebook Messenger.

Tencent may have beaten Facebook to the punch, but the figure above is one of the Messenger chatbot stats that depicts how busy the app’s developer community has been since 2016.

20. 84% of messages on Messenger are opened.

Once enough marketers appreciate this convenient fact, Facebook Messenger chatbot stats will soar even further. The interactive nature of AI alone is enough indication that opened messages would translate to higher customer engagement.

21. In January 2020, there were more than 750 bots on Slack.

This number is not as impressive as the hottest Messenger chatbot statistics, but some bots on Slack can be integrated with other platforms such as Kik, Skype, and Viber as well as company websites.

22. Around the same time, there were over 500 bots on Telegram.

While this figure will not match the latest Facebook chatbot statistics soon, the app has great ambition.

Telegram targets to have one billion users by 2022, which is a gargantuan task since just 200 million people actively used this messaging service monthly in 2019.

The Chatbot Industry

23. The size of the global chatbot market is expected to grow from $2.6 billion in 2019 to $9.4 billion in 2024.

Some industry observers have forecast a compound annual growth rate of 29.7% during this period. The increasing need for customer assistance 24/7 would fuel such strong, steady growth, so customer support chatbot statistics should vastly improve, too.

24. 58% of corporate chatbot adopters are B2B.

In comparison, the latest chatbot market statistics say that just 42% of B2C companies have embraced artificially intelligent programs.

25. 80% of business leaders from the Netherlands, the United Kingdom, France, and South Africa have deployed or planned to adopt chatbots and other automation innovations by 2020.

This was based on a 2016 survey. If all of these decision-makers fully implement AI-powered bots before 2021, the explosion of chatbot growth statistics will be on the horizon.

26. Bots can answer up to 80% of routine questions encountered by businesses.

So, what is a customer service chatbot? It is not yet a full-fledged replacement for a human customer-care agent. But an AI representative already drove down customer service costs by 30% by 2017.

27. Giving 49% of customer service roles to bots is estimated to translate to $23 billion in annual salary savings.

If other workforce-related costs like health insurance are taken into account, rosier chatbot savings statistics are expected.

28. 55% of consumers like to engage with bots when dealing with businesses online.

This stat confirms that people trust chatbots.

29. Americans account for 36% of all chatbot users worldwide.

One of the most sought-after chatbot industry statistics right now, this number suggests which nation will dictate global bot adoption.

Indians and Germans were the second- and third-biggest fans of these AI-driven programs, representing 11% and 4% of all users, respectively.

30. In 2020, 50% of search queries will be made via a voice assistant.

This estimate has got to be among the top chatbot marketing statistics SEO professionals should crunch from here on out.

31. Financial institutions would be able to automate 9 out of every 10 client interactions through AI-driven bots in 2022.

Apart from banking, food, retail, health care, and telecommunications are the leading industries with the most promising chatbot implementation statistics.

For instance, a 2017 UK study predicted that the interactions between healthcare patients and bots would reach more than 75% two years from now.

32. The use of AI-powered bots in sales could increase conversion rates by 30%.

It is one of the many chatbot sales statistics that demonstrates how effective machines could be in engaging prospects and leads.

33. When it comes to business uses, 46% of companies turned to chatbots for voice-to-text dictation in 2018.

What’s more, 26% of enterprises relied on AI-powered bots to improve team collaboration. On the other hand, 24% used intelligent assistants for employee calendar management.

And these HR chatbot statistics are just the tip of the iceberg. More viable use cases should become more popular soon, as these programs gain mainstream adoption.

34. 61% of executives believed that conversational bots would boost employee productivity by following up on scheduled tasks automatically.

60% thought these intelligent programs would enable employees to handle client queries. Additionally, 57% envisioned the delivery of personalized attention to website visitors.

However, these AI chatbot stats fail to shed light on the four common classes of conversational bots. These are informational, transactional, enterprise productivity, and device control.

35. At least 64% of all age groups said they would use chatbots for specific purposes.

Furthermore, only 4.67% of online shoppers 30 to 44 years old said they do not like to use bots. Just 5.09% of consumers aged between 18 and 29 as well as 6.70% of shoppers 45 to 60 years old shared the same sentiment.

These chatbot statistics in 2019 clearly showed that internet users are warming to the idea of using pieces of intelligent software.

36. 47% of American millennials noticed that chatbots do not always give helpful answers.

Despite this, the chatbot benefits statistics are not lost on most young adults in the US. 67% of these consumers said they would purchase a product or pay for a service from brands that implement an AI-driven bot.

37. 52% of consumers would give up mobile apps and websites in favor of voice assistants to experience the convenience and be able to do multitasking.

Just 35% of people said voice assistants seem like real people. But the fact is more than half of consumers now prefer speaking to robots instead of browsing the Web themselves. This should make the figure above among the most compelling AI chatbot statistics collected.

38. 41% of consumers would choose a voice assistant over a mobile app or a website to automate routine shopping activities.

This has got to be one of the most telling ecommerce chatbot statistics ever reported. Bots have not made apps obsolete yet. But these AI-based programs are already becoming a part of the lives of many contemporary consumers.

39. In 2017, 82% of users spoke to a voice assistant to seek information.

Most of these people do so to learn about the headlines, ask for recipes, and receive relationship advice. Only 28% of them have instructed a bot to make a payment or send someone some money, though.

These statistics about chatbot emphasize which functions tech-savvy consumers are willing to delegate to machines.

40. About 52% of US consumers want to use voice assistants for healthcare-related activities.

The latest chatbot uses statistics revealed that voice assistants have been hardly implemented in healthcare. The survey showed that just 7.5% of people have done so.

Although the youth are more likely to talk with voice assistants, people aged 45 to 60 are keener on the idea. Now that the word is out, healthcare chatbot statistics are highly likely to impress in the near future.

41. In 2019, 48.78% of women from all age groups who shop online considered chatbots as their preferred channel of communication.

In comparison, just 36.81% of men felt the same. More than 41% of male consumers still chose email. These chatbot stats mean that there may be some correlation between gender and bot adoption.

42. 47.66% of Android users selected chatbots as their primary channel of communication when shopping on the internet.

On the contrary, just 41.91% of iOS users had a similar mindset. Nevertheless, these chatbot stats in 2019 are proof that email correspondence is losing its appeal.

The Politics of Chatbots

43. By 2018, Cortana was the most popular artificially intelligent assistant in the workplace, with a 49% usage rate.

Microsoft might have dominated the work-related chatbot statistics, but the company’s lead was slim. Apple’s Siri was used by 47% of professionals, while 23% of companies have employed Google Assistant.

Artificial intelligence statistics indicate that only 13% of businesses adopted Amazon’s Alexa.

44. Alexa-powered smart speakers accounted for 70% of the US market share in 2019.

The chatbot marketing stats of Google Home and Apple HomePod pale in comparison. The former controlled just 25% while the latter managed to maintain 5% only.

45. In 2018, 26.2% of US adults owned a smart speaker.

Moreover, one-fifth of all UK households had a smart speaker by June 2019. Germany’s household penetration rate reached 11.6%, while Ireland and France had 9.6% and 7.4%, respectively.

These important 2018 chatbot market facts must be music to the ears of Jeff Bezos and other Amazon executives.

Key Takeaways

It is too early to declare AI-powered software as a force for good or an existential threat to civilization at this point in time.

Instead of worrying about the bots themselves, many of the most recent chatbot statistics and facts suggest it is humans who create intelligent programs that must be kept in check. After all, artificial intelligence is not natural.

Frequently Asked Questions