Since Facebook launched its annual F8 conference for robot developers in 2016, Microsoft has followed closely, with the hype, excitement and speculation surrounding chatbots. Microsoft CEO Satya Nadella has said that chatbots will “fundamentally change the way people experience computing” and change the way content and services are created and consumed on the web.
On the Messenger network alone, there are currently more than 100,000 robots and developers. More than 10,000 businesses are already using or developing their own bots, and by 2020, 80% want to have chatbots for interacting with customers.
This article will explore what chatbots are, why their popularity suddenly soars, how they significantly change the way we browse and interact with the web, why web developers and designers should focus on this paradigm shift, and how they What can be done specifically to keep up with the trend.
Key points
What is a chatbot?
The following is a Google trend chart showing the exponential growth in people's interest in "chatbot" search:
By definition, a chatbot is a computer program powered by machine learning and natural language processing (NLP) that anyone can interact with to complete tasks or find information and services. They are primarily found on instant messaging platforms such as Facebook Messenger, Kik, Telegram, Slack, Skype and WeChat, which have existing user bases.
Chatbots are intelligent, automated conversational identities that make clicking the drop-down menu button, selecting search functions, and navigating on website pages redundant. They provide instant, specific results in a smooth conversation format.
How will robots change network interaction?
Current web interactions usually involve extensive Google searches, getting a bunch of search results, browsing one by one until we find what we want, and finally taking some action.
On the other hand, using robots, browsing and navigation elements will be replaced by simple question-and-answer conversations.
To better understand how online interactions will change, let's take any e-commerce website as an example. Your ultimate goal: You want to buy a beautiful blue shirt for the meeting, but don’t want to spend more than $30.
Normally, you might visit an e-commerce website (assuming it is filled withclothes.com): you hover over the category at the top, select a men's/women's shirt, and then be directed to another login page… …This is even less than halfway.
Now on the product list page, you either set some filters to suit your budget and color preferences, or you will skip this step and start browsing your shirt right away.
Here, you need time and patience to browse all the options and find the options you like. You will check the size chart, maybe, maybe decide to buy. After that, checkout and payment will guide you through a series of actions until you finally complete the order.
Now imagine that the same e-commerce website decides to get a chatbot (like countless other brands like H&M, Tommy Hilfiger, Burberry and Sephora). Let's call it Clothes Bot. Ideally, they would place the bots where their audience is, which is likely Facebook Messenger or their website login page.
The following is a simulation of interacting with Clothes Bot:
As a user, you can talk to this bot to find products directly from Facebook chats, almost like you send a message to a friend.
Before I explain in depth how developers and designers can participate in the robotics revolution, I will first introduce why they should care about this issue first and what changes will happen in terms of development and design.
At present, all major companies such as Google, Microsoft, Facebook, IBM and Amazon have released many open source robot building tools and frameworks to help developers create robots. They have also formed an AI initiative with companies such as Amazon, Twitter, Apple, Intel and Baidu to address ethical and privacy issues that may arise during the development and scaling phase.
1.2 billion people use Facebook Messenger every day and introduce robots in the same window to help businesses interact, market, and sell with customers, with a larger scope than when the app first appeared.
Instead of downloading the app, users can use a summary app like Facebook, Kik, Telegram, or any other standalone chat-based platform where countless bots are available for people to search and talk.
Kik hosts a robot store with a variety of robots for their user base to search and talk to. Slack provides an excellent list of robots for all its corporate accounts to increase productivity and complete tasks faster. This list continues.
Essentially, the average web user simply searches for the brand/company/service from their native messaging app or the web, and starts chatting with their bots like they would with friends on Messenger.
Early adopted adopters in the field of technology are supporting the robotics revolution. It's worth considering what this means for developers and designers. What will the network look like in the next few years and how will robots shape it?
From mobile apps to addictive social networks, "Web 2.0" has witnessed some iconic developments such as video sharing websites (YouTube), blogs, wikis and RSS feeds (Feedly), collaborative consumer platforms (Craigslist) , Uber). However, at present, two cross-trends make the rise of chatbots the next possible stage of the network:
Mobile app fatigue. A large number of smartphone users download very few or no apps per month. And with over 3 million apps available in both app stores, it’s becoming increasingly difficult for businesses to build unique apps, and even harder to stand out from the noise.
Popularization of instant messaging platforms. For the first time in history, the number of monthly active users (MAUs) of the four instant messaging apps exceeds that of the four social networking platforms:
Users obviously prefer instant and interactive interfaces. It makes sense to be active on a chat-based platform for businesses that want to target customers in their locations. Instead of asking customers to jump from one login page to another, integrate all the basic features into a chat environment.
The following is an overview of how conversations as a platform will change the different elements of the end user interface:
Design meaning
2017 Web Applications
Web Applications in 2020
Development Meaning
2017 Web Applications
Web Applications in 2020
The hybrid interface combines the best features of CUI and GUI into one interface—instant, personalized conversations from conversation modules, and rich UX from graphics modules.
For example, look at WeChat. With about 800 million monthly active users, WeChat has merged a messaging portal and small applications, from e-commerce to social networks, into a chat window.
In short, people can use it not only to chat with people and businesses, but also to do things like booking cars, booking movie tickets, playing interactive games, watching news, getting stock updates, and more without leaving chatting.
What we are seeing now is conversations as a platform, brands and businesses are creatively trying to use robots to develop marketing, sales, engagement and support strategies. For example, Tommy Hilfiger, H&M and Sephora are using chatbots on instant messaging software (Facebook, Kik and Telegram) to initiate fashion-related conversations and recommend products based on user preferences and dislikes. Businesses like Uber, Domino and 1800-Flowers are receiving orders and payments via chat (Facebook), while Netflix, TechCrunch and Casper are using bots to engage their fans in easy conversations, providing movie suggestions, news updates and useful insights— All of this is based on user behavior.
The following is an example of interaction with the Wholefoods robot, which abandons traditional menus with graphic buttons and Q&A:
More specifically, each robot response may be simple text or a micro application itself. This gives developers and designers the opportunity to create a wealth of cards that display text, images, product carousels, payment gateways, two-player games, music players, and more—all.
In view of the limitations of chat windows, developers need not only focus on building infrastructure, but also on the overall experience.
With the navigation menu and many login pages or screens, websites and applications lag far behind in providing specific answers. Search fields? Absolutely not enough!
On the other hand, the robot will use super related cards that provide specific answers to answer questions. For example, instead of using a music app or website that needs to jump between screens, you can tell the bot what artist or genre you want to listen to, and the bot will return to super related tracks for playback.
A better use case would be e-commerce sites where browsing product pages is definitely a customer challenge and can be solved with a simple chatbot.
In the robot world, asking a chatbot “Do you have any cool sunglasses?” may trigger the bot to ask the user’s counter-question, such as “What occasion do you want to use it?” or “Did you like it in particular?” or “Did you particularly like it?” The color? ”
According to the user's answer, the robot may return a list of texts with available sunglasses, or return a rich card with a carousel to scroll through the results, each with a purchase button that immediately triggers the payment gateway.
The following is an example of interacting with the Tommy Hilfiger robot that guides the user into the conversation path:
Creating a bot from scratch is fundamentally different from creating a website or an application.
If you want to catch this wave, new developers and designers who are new to this field will have to change their workflow, tool sets and holistic approaches.
First, they will no longer have blank tables with boilerplate code. Instead, they will have to design within the chat window and focus on infrastructure and end-user expectations.
Robot frameworks, model tools, robot prototyping tools, test environments, conversation process designers, and analysis will be key (described below).
To learn what a chatbot platform looks like, get a free and powerful platform like Bottr (created by myself), which allows you to embed your own chatbot widgets into your login page to Visitor interaction.
Let's see how designers and developers stay ahead of their position...
While developers have the option to write robot code from scratch, the robot framework provides a canvas that reduces the need for developers to write by allowing them to access the robot builder SDK, robot connectors, developer portals, robot directories and emulators the amount of code.
They allow developers to assign roles, intents, retrieval patterns, response sets, etc. to the robots they create. The most popular frameworks include:
Microsoft Bot Framework. A great tool that contains two modules: robot connector and NLU component LUIS (Language Understanding Intelligent Service).
This simple method of doing it will guide you to develop and deploy robots for Facebook Messenger using node.js.
Robot frameworks, tools and platforms. (Source)
Api.ai. Built based on core elements such as agents, entities, intents, operations, and context. Read this informative library to learn about key concepts, or check out SitePoint's popular article "How to Build Your Own AI Assistant with Api.ai".
Wit.ai. Create a bot by streaming the sample conversation to its knowledge base. Learn how CreativeWorkline uses wit.ai to create bots in response to incoming Messenger chats.
Bottr.me. An end-to-end solution and chatbot creation tool that anyone can create a bot in one interface, train a bot to make it smarter (through conversations), deploy it to an existing social network and get analysis results of a bot conversation .
When creating intelligent robots from scratch, you must have an in-depth understanding of various AI concepts and best practices. The core of machine learning lies in the ability to learn, recognize patterns, understand natural language and respond accordingly.
Designing intelligent NLP chatbots requires you to understand and implement tokenization, entity recognition, normalization, voice tagging, dependency analysis, intent and sentiment analysis.
As a robot developer, you will need to understand the concept of neural networks, information retrieval models, type of machine learning algorithms, sequence-to-sequence architecture, submodals, perceived locations, anchors, metamodels more than ever before. etc.
Compared to websites that rely on graphic elements, the conversational interface relies on smooth conversations—scripts and dialogue flows rather than graphics.
Making robot friendly, engaging and intelligent (rather than stiff, difficult to understand or frustrating) is a real challenge for developers.
As a robot developer, you must learn to classify the intentions and emotions behind nonlinear interpersonal interactions. Verifying user input on the CUI and classifying it is more complex than a standard GUI, and classifiers are definitely a priority.
The better your chatbot is able to distinguish input from different users, the smarter the conversation it will have.
I recommend you read these important conversation design tips before you start.
In "Conversational UI Principles - The Complete Process for Designing Website Chatbots", Leszek Zawadski, founder of UX agency TheRectangles, details the best conversational UI principles and how to put them into practice.
Robots are new, robots are fresh, and they are just beginning to gain crazy attention.
Given that AI and robots are still in their infancy, it may be difficult to find market-wide statistics, although raw data from different platforms provides detailed reports on their monthly active users and participation rates. There are some chatbot stores and catalogs to get inspiration from. There are thousands of businesses and brands with chatbots that can be talked to.
Before starting to make robots, it is best to study how users interact with them by talking to them yourself.
By default, chatbots need to be built and optimized for chat windows, rather than for websites or applications built on blank HTML and CSS-based forms.
Therefore, it might be a good idea for designers to test their ideas using robot prototyping and modeling tools before they actually implement them.
Some of these tools are BotSociety.io, Botframe, and Walkie, which allow designers to easily create and test user processes for conversations before actually deploying the bot.
For websites and mobile applications, you may be accustomed to considering participation metrics such as average session duration, bounce rate, click-through rate, dwell time to determine what works and what is not.
For robots, you must consider the number of messages received by the robot, the number of conversations, chat duration, chat scores, conversation funnels, intention analysis, sentiment analysis, user life cycle and other factors.
It will no longer be about monthly active users; the effectiveness and relevance of each bot session will be important.
If you use frameworks such as Watson or wit.ai to build a bot and deploy it to a third-party end-user platform (any messaging platform), robot analytics tools such as Dashbot and Botanalytics are for in-depth engagement and Retention rate reporting is very useful.
On the other hand, there are certainly advantages to using end-to-end solutions such as Bottr. It provides detailed insights and notifications to tell you how people interact with your bot.
The robot revolution has arrived, and only time can tell how far it will go. The chatbot community, conferences, discussions, and open source networks are growing, and a large number of developers join this ranks every day.
All robotic frameworks come with their own discussion communities, such as IBM's Watson Developer Cloud.
Many conferences explore artificial intelligence and robots in more depth. You can find the full list of 2017 here.
Communities such as Botness, UXDesign, Slack Developer Gathering and Open Chatbots are perfect for developers, designers and enthusiasts to come together to share ideas about AI and robots.
BootCube provides a comprehensive list of robot podcasts, prototyping tools, platforms, SDKs and more on GitHub.
Get inspiration from this tutorial to create a simple Twitter bot in Node.js.
Learn how Adrian Zumbrunnen takes a different direction by integrating automated chat into his website.
To get inspiration from other developers, you have to read how Joyce Echesssa builds Facebook bots with Node and Heroku in response to movie queries, and how Android Advance makes resume bots using Python and Watson frameworks.
Hopefully, after reading this article, you have a better understanding of how as a web developer and designer builds robots and become part of the AI revolution.
So, start acting, roll up your sleeves, and start creating robots now!
Want to learn more about chatbots? Please check the following link:
View SitePoint Premium for more books, courses and free screen recordings.
Successful chatbot user interface (UI) should be intuitive, user-friendly and engaging. It should have a clear and concise design that allows users to browse and interact with chatbots easily. Chatbots should also provide fast and accurate responses and be able to handle complex conversations. Personalization is another key element as it allows chatbots to provide a more tailored experience for each user.
Chatbots can significantly enhance the user experience of a website by providing instant customer support, answering FAQs, guiding users through the website, and even helping with purchases or bookings. They can also provide personalized recommendations based on user behavior and preferences, which can improve user engagement and conversion rates.
Chatbots can bring many benefits to businesses. They can handle customer consultations 24/7, reducing the need for manual customer service representatives. They can also automate repetitive tasks, freeing up time for employees to focus on more complex tasks. Additionally, chatbots can provide valuable insights about customer behavior and preferences that can be used to improve products and services.
There are several platforms and tools that can help you build chatbots for your website, such as Chatbot builder for Hubspot, and Landbot.io. These platforms provide easy-to-use interfaces and pre-built templates that can help you create chatbots without any coding knowledge. However, if you have specific requirements or want a more custom chatbot, you may need to hire a web developer or chatbot development company.
The future of chatbot development is expected to be driven by advances in artificial intelligence and machine learning. This will enable chatbots to understand and respond to complex queries more accurately. Personalization will also become more complex, and chatbots will be able to provide a more tailored experience based on user behavior and preferences. Furthermore, we can expect chatbots to be further integrated with other technologies such as virtual reality and augmented reality.
Chatbots can significantly improve customer service by providing instant response to customer consultations, thereby reducing waiting time. They can also handle multiple consultations simultaneously, ensuring no clients need to wait. In addition, chatbots can provide personalized customer service based on previous interactions, thereby improving customer satisfaction and loyalty.
While chatbots can automate many customer service tasks, they cannot completely replace human customer service representatives. Chatbots are very good at handling simple and repetitive tasks, but they still struggle to handle complex queries and situations that require empathy and emotional understanding. Therefore, the combination of chatbots and human representatives can provide the most effective customer service.
To make your chatbot more attractive, you can use conversation language, add a little sense of humor, and personalize the chatbot's response based on user behavior and preferences. You can also use rich media such as images and videos to make the conversation more interactive. Additionally, you can constantly update and improve your chatbot based on user feedback and behavior.
Some of the challenges in chatbot development include understanding user intentions, handling complex conversations, and providing accurate responses. Developing chatbots that understand and respond to different languages and dialects can also be challenging. In addition, ensuring the privacy and security of user data is a major issue in chatbot development.
The success of a chatbot can be measured using various metrics, such as user engagement, conversion rate, customer satisfaction, and the number of tasks successfully completed by a chatbot. You can also use user feedback and comments to evaluate the performance of a chatbot. Regular monitoring and analysis of these metrics can help you identify areas of improvement and ensure the success of your chatbot.
All picture formats are .webp.
The above is the detailed content of Chat Bots and the Future of Web Development. For more information, please follow other related articles on the PHP Chinese website!
![[Web front-end] Node.js quick start](https://img.php.cn/upload/course/000/000/067/662b5d34ba7c0227.png)
