Facebook Messenger is Now Friends with Artificial Intelligence

Facebook has been following the biggest online trends since its debut. By this, I mean they keep evolving and they stay dynamic within the social media realm. If you think about it, how many times have you seen new elements appear on your Facebook feed? I remember when I first got Facebook about 10 years ago. The Facebook of then and the Facebook of now have pretty much nothing in common apart from the fact that it serves as a social platform for millions of people to share their lives on and to communicate. One of the main features of Facebook that keeps improving is Facebook Messenger.

Last year, Mark Zuckerberg announced at the annual F8 conference that the Facebook Messenger app was the fastest growing app in the US, with up to 900 million monthly active users. Between Messenger & Whatsapp people were sending up to 60 billion messages per day worldwide. The global SMS volume alone was 20 billion per day. So Facebook processed about 75% of the world’s mobile messages and that with a startling increasing growth rate.

With a user base like that, Facebook decided to take the Facebook Messenger app just one step further. The goal was for it not to be just another app but an app that could help you with queries, like order flowers online (just like the image below) without the need of any human input.

But what effect would it have on the digital world?

Before answering this question, let me dive a bit more into the topic of Facebook and Artificial Intelligence. After covering this, we will take a look at where Facebook Messenger is now, and why the new AI system Facebook launched in 2016 is not as much of a success as anticipated.

Facebook Messenger: a smart platform

In April 2016, Facebook came out with a big announcement: the Messenger Platform was launching. It was meant to be Facebook Messenger on steroids: it included bots that could order pizza for you whenever you wanted it to, give you the weather everyday, send your location to your friends and more. But you could do this by just having a conversation with your Facebook Messenger. It may seem quite simple at first, but it got me, and the rest of the digital world, pretty excited. 

The goal behind Facebook Messenger

Facebook has always been about satisfying the customer. The main goal behind the new Facebook Messenger was to create a more engaging customer experience that was more contextual, conversational, and convenient whenever people wanted to buy or find something.

To do that, Facebook teamed up with a few businesses while still making sure that its users were in control. What does that mean? Let me give you a simple example. If you wanted to buy flowers, you simply had to login to your Facebook Messenger, find a bot from a flower shop,  and chat with it. Tell him what kind of flowers you wanted, the colour, the price and the bot would find the perfect bouquet for you. From there, you just had to click on a link that would redirect you to a webpage with your order, and pay.

For companies, this meant that customers were able to communicate to the bot version of their preferred brand on Messenger. With the help of bots, customers could now do things like purchase products, ask questions about the services, and request information without having to call or speak to anyone!

The goal from there was for the platform to learn more about the user’s interest over time. From that, businesses could feed more relevant data to users. So after buying flowers, the Facebook Messenger bots could contact you again on Valentine’s Day to know if you were interested in buying another bouquet of red roses for example or, if you preferred a bouquet of yellow flowers as it seemed like you liked the yellow colour a lot.

So how does it all work? Let me explain what happens behind the screens.

The logic behind it

Businesses can create their own Facebook bots using an API to send and receive request. So for example, it would receive a request from you, saying that you want to buy flowers. The Facebook Messenger bot would then send that information to the flower shop and receive back information about what kinds of flowers or bouquets are available.  The bot would then give you the best information based on what your query was.

Think of an API (application programming interface) as a list of commands that one program can send to another. I like to imagine an API as a bridge between two softwares. In this case, the bridge would be between Facebook and the flower shop website for example.

The API allows bots to receive the information and display it with images, and links to make the customer experience as enjoyable as possible. Through these bots, you were able to buy flowers, book flights, or even buy clothes. The Facebook bots’ job was to help you find what you were looking for and, once you were ready to purchase, it would redirect you to a web page to pay.

So how were people able to know what they could buy or which business was linked to the new Facebook Messenger? Through a search bar in Facebook Messenger, people could type the name of a company and, if it appeared, could chat with the company bot.

Facebook was not the only one developing Artificial Intelligence. WeChat, for example, has been very successful in Asia with implementing AI. But what does this mean for businesses? What were the hopes and fears that the new Facebook Messenger brought to businesses when it launched? Let’s take a look.

Facebook Messenger, Artificial Intelligence, and its possible impact on business

Powered by artificial intelligence and through natural language services, Facebook Messenger allowed users to interact directly with brands and services. So your customers could engage directly with your brand via conversations within Messenger with customised bots which you could build via the Facebook open platform. But what does this new way of communication mean for businesses?

When the new Facebook Messenger was launched, businesses quickly understood that it could mean a few things. It could:

• Make humans obsolete
• Impact mobile apps
• Change the purchasing pattern

Let’s take a look at each of these in greater detail.

Will Facebook Messenger make humans obsolete?

One of the biggest fears that people have when it comes to technology is that they will get replaced. How many times have you heard someone say “this machine will replace humans” or “it will take our jobs away”?

Facebook Messenger was able to provide personalised communication as well as a “real” person could. Moreover, it could replace phone conversations. A lot of people, nowadays, don’t like to make calls, myself included. I would much rather book a flight online for example. The Facebook Messenger bots could prevent people from having to make countless phone calls, and as a result it would also reduce hold times since less people would be calling real customer service representatives. These customer service representatives, more often than not, can only help one customer at a time while a bot has no limit.

So what happens when there is no need for human customer support anymore?

There is a very debatable topic and, while customer support jobs may be in danger, it could also create demand and generate new jobs for those who know how to script and implement bots.

But humans might not be the only ones losing their jobs. Mobile apps might become less popular, or even obsolete, as a result of Facebook Messenger bots. How and why? Let me explain further.

Impact on mobile apps

Mobile apps have been on the increase in the last few years. More and more businesses are investing in them. They represent a new way to advertise and sell their products to people who are always on the go. However, these same people would basically be able to find everything they want through the new Facebook Messenger.

A simple stand alone app will be outclassed by Facebook Messenger due to the web of information it has access to and the kind of adaptedness the AI presents. It makes the user’s experience much more personalised and “human”. People would feel like they are conversing with someone that can give them an answer to the specific question asked, and how they asked it, rather than just scrolling and choosing from a set of questions and answers in an app. The bots would also have details such as your id, and the history of your past conversations.

So if you put yourself in the shoes of a customer, what would you prefer?

Change the purchasing pattern

Another way in which Facebook Messenger could affect businesses is that it could change the purchasing pattern of consumers. Today, when people look for a product to buy, they will go through the awareness stage (a stage where they will search for you online and get to know your brand and your competition).

After this, prospects would then refine their choices – choose between you and your competitors offerings, before finally purchasing. Throughout this sale cycle, your prospect would evaluate you and your competitor’s content then choose the best customer experience etc.

Now, imagine if you are looking for a product and ask your Facebook Messenger bot to help you out. Instead of going through the awareness stage, your bot would take you directly to a specific brand where you could then buy the product you were looking for. Instead of going through content and making your own choice, bots could make you skip the awareness stage and take you directly to the purchasing stage. Chances are you will like what the bots present to you since they know what your preferences are and what you value. So how do you think this will change the way businesses market to consumers?

No matter what you think, it seemed like Facebook Messenger is out to change the business world. It’s almost been a year since it launched. So where is it at? How much has it evolved since then? Let’s take a look at what happened since April 2016.

Facebook Messenger and Artificial Intelligence in 2017

Facebook was already using AI to track users, scan their profiles and ultimately, deliver targeted ads. So Facebook didn’t start AI with Facebook Messenger. It actually began to really focus on AI back in 2013 when they developed “FAIR” (The Facebook Artificial Intelligence Research). However, the Facebook Messenger bots were suppose to take it one step further. But it didn’t and something happened.

Facebook Messenger: an AI fail

While Mark Zuckerberg had great visions for Facebook adopting AI, things didn’t go quite like he had planned. And in February 2017, Facebook was starting to scale back after its bots hit a 70% failure rate.

So what happened? How come Facebook could not achieve WeChat’s success?

From the get go, it seemed like the Facebook bots were only able to fulfill 30% of the requests without needing human input. For a system that is suppose to be independent from humans, this was too low.

The Facebook bots were not, for the most part, built by Facebook developers. They were bots developed by businesses on the Facebook platform who offered a link between Facebook and their brands. While Facebook had played with AI a bit, most developers from other businesses hadn’t. Facebook began to realise that the problem with these bots was that most of them had major errors and were not able to understand most human requests. This ended up limiting users and they were only able to use very specific sentences or specific requests.

Should Facebook have seen it coming?

Some believe that it would have been easy for Facebook to see it coming. The AI technology isn’t that well developed and from what we know, it already caused a few hiccups for a few companies.

Facebook could have taken a look at what happened to their main competitor, Microsoft, when it tried to launch the Tay chatbot in March 2016. Within a day, Tay went a bit crazy and started tweeting racist comments. Microsoft had no choice but to stop and apologise for it.

Moreover, even if Facebook did toy with AI before launching the new Facebook Messenger bots, did it take the time to develop it enough? If you closely look at your news feed for example, you can see that the AI struggles to discern fake news.

It seems that Facebook rushed the release and just didn’t have enough experience in AI. An almost typical “first mover” scenario. However, Facebook hasn’t lost hope. With more experience comes more knowledge and I wouldn’t be surprised to see the Facebook Messenger bots comeback in a few years, when Facebook is truly ready.

The future of AI and Facebook

It seems like it will take a little longer for Facebook to fully understand and correctly implement AI. But it isn’t all bad. This lesson has helped Facebook realise that they need to review their objectives and that they do not have to have the first fully functioning AI. Rather, they must have a functioning AI when they implement it. Therefore, they are now focusing on training Facebook Messenger bots to understand and correctly interpret human requests and making sure that they meet the quality standards set by Facebook.

You need to know that the world’s biggest tech companies are obsessed with Artificial Intelligence. Facebook is just beginning to scratch the surface with this latest feature. But if it succeeds in implementing it correctly, it could allow businesses to extend their reach and communication into personal mobile devices via online chat, without any real people on the other side. It represents a breakthrough for all consumer-based industries that need to scale or engage in building brand loyalty.

What do you think is next for Facebook and the AI world?