For decades, the job of a UX designer came down to designing graphical user interfaces (GUIs) that allowed end users to get things done. With the first GUIs appearing in the 1980s, the design of business applications was limited to spreading icons, menus, radio buttons and check boxes more or less logically across the screen. Luckily, these times are over for good.
Modern business applications are on par with the best consumer applications. Their GUIs are the result of a design-led development process: end users have been observed and interviewed, wireframes and prototypes have been validated and iterated with them, and visual designers have put their best foot forward in creating beautiful user interfaces. And by using interaction patterns like tap, swipe and pinch instead of type, click and scroll, GUIs are more intuitive than ever before. Putting it in context of SAP: SAP Fiori is the living example of how the graphical user interface of SAP has transformed in just a couple of years.But…. all good things come to an end. Don’t worry, I don’t mean to say that SAP Fiori comes to an end. It has a firm seat on our roadmap for the years to come (actually, it’s the key user experience of SAP software). Neither do I intend to say that GUIs will cease to exist anytime soon. However, the message I would like to convey here is that we are starting to reach the limit of graphical user interfaces.
Now you might wonder: what’s next? Is there a better way to interact with products and systems around us than a GUI? What could possibly replace all those websites and applications that we use every day? Before answering these questions, let’s first look at some trends in the market.
Trend: Artificial intelligence has entered everyday life
Artificial intelligence (AI) gives systems the ability to understand, learn, predict, adapt and act autonomously. Recent research by experts from Yale and Oxford University predicts a 50% chance that artificial intelligence will outperform humans in every job in 45 years’ time. Already in 2014, the internet documentary Humans Need Not Apply made us aware about the rapid pace in which robots are taking over human labor.
While there is ongoing debate whether robots will ever be smarter and better than humans, AI is already present in many forms today – more than the average user realizes. It has an impact on the user experience of applications like Spotify and Netflix, where it provides recommendations based on previous choices. It learns our behavior and habits, like Nest’s thermostat. It drives cars autonomously through traffic. It recognizes objects in pictures, and it detects emotions when we look at billboards. It makes proactive recommendations when we shop online. It curates content on social media. And so on.
While many people still associate AI with science fiction, reality is that many products and services already use AI to offer a better user experience!
Trend: Messaging applications are becoming the new mobile home page
Messaging applications like WeChat, Facebook Messenger and WhatsApp have shown a strong user growth over the last years. They have evolved from simple social conversation tools to self-expression platforms by introducing emojis, picture & video sharing, filters, and special effects. And they have an enormous reach: mobile messaging applications have over 3 billion active users globally!
Especially for millennials, messaging-based applications are becoming the second mobile home screen. And they do not only use messaging apps to stay in contact with their friends. Messaging is also their preferred way of interacting with businesses, for example for ordering stuff, making reservations, enquiries or complaints. Using messaging apps, these interactions are conversational in nature and organized in threads. They remember the user’s identity, preferences and context.
Let’s look at Facebook Messenger. More than 2 billion messages are sent every month between people and businesses, both automated and people-initiated. By using bots, Messenger offers business to reach people at scale but interact with each of them individually. Facebook introduced an API for building chatbots for Messenger in April 2016. Today, they have over 100,000 monthly active bots, and the number is growing.
Another example is WeChat. WeChat has clearly been leading the trend of business messengers, offering official business accounts since 2012 and WeChat Pay since 2013. It started as a simple chat app, but has grown into a full-blown messenger offering a mixture of messaging, gaming and social media. Using WeChat, people can order and pay for food, pay their bills, play games, keep in touch with friends and track deliveries. Its 900+ million active users (mostly Chinese) are using WeChat when they wake up, and when they go to bed – and every moment inbetween.
And since a few weeks also WhatsApp, the largest messaging platform globally, embraces the potential of business messenging. It has introduced a pilot program for verified business accounts, underlining the trend of messaging applications becoming the new mobile home page.
Trend: Voice usage is growing
Until a few years ago, using voice commands to interact with applications wasn’t very common. Some businesses operated interactive voice response (IVR) systems to streamline call center operations, or to offer services outside business hours. The experience wasn’t always good. Cars started to feature voice control, like entering a destination in the satnav system, or dialing a contact without taking your hands off the wheel. And some might have tried speech-to-text recognition software for writing emails or documents. To put it short: voice interfaces haven’t had their breakthrough moment yet.
But that is going to change! The benefits of using a voice interface compared to a GUI are obvious:
- It’s faster. The average human can type 40 words per minute, but speak 150 words in the same time.
- It’s easier. No need to operate a device like a smartphone, keyboard or mouse. You have hands-free interaction so you can keep doing what you need to do.
- It’s cooler. At least for now, while voice interfaces are not so ubiquitous yet. Won’t you admit experiencing a little moment of joy when the system you’re talking to, actually understands you?
The introduction of Apple’s Siri (in 2011), and Amazon’s Alexa (in 2014) have accelerated the adoption of voice as a computing interface by consumers. Machine learning has greatly improved speech recognition accuracy (to over 90%), and reduced latency i.e. the time it takes the system to respond (to less than 2 seconds). Search platforms like Google and Baidu have seen a tremendous increase in voice queries during the last couple of years.
Where the average two-year old was able to swipe an iPad, the same is happening now with Apple Siri, Google Home or Amazon Alexa which are easily commanded by young children to play their favorite music, or to turn on the lights. And anyone with children at home will acknowledge that Generation Z has no reservations in talking to a machine – something that many older people might still feel uncomfortable about!
From “mobile-first” to “conversational UI-first”
All these trends have great impact on UX design. Do you remember recommending specific mobile apps to your friends a couple of years ago, thrilled as you were with yet another masterpiece of design? No one is doing that anymore. Mobile apps are a commodity, you simply expect every company or service to have one. The average user might have 30 – 40 apps installed, but uses only a handful daily. And what’s worse, people don’t want to go anymore through the burden of finding yet another app, downloading it, logging in or creating a profile, and learning how to use it. They rather prefer apps they already use and know, and which remember their preferences and past interactions. This is where conversational user interfaces and chatbots come into play!
The growing usage of voice, the rise of mobile messaging applications and today’s powerful artificial intelligence are all signs of a platform paradigm shift happening right now. As Gartner and others have pointed out, roughly every decade a new platform paradigm emerges that works its way through the industry. As a result, human-to-machine interaction is also changing.According to Gartner, conversational AI platforms (CAPs) will be the next big paradigm shift in information technology. They predict that in 2020, the average person will have more conversations with bots than with their spouse. Conversational systems are clearly a technology trend: Accenture’s technology vision foresees that in seven years’ time most interfaces will not have a screen anymore but will be integrated into daily tasks. Artificial intelligence will become a digital spokesperson for companies.
All of this means that we will see a shift from “mobile-first” to “conversational UI-first” in application design. Sooner rather than later, companies need to design for an improved customer experience by offering intelligent assistants and chatbots.
And the winners are …. the users!
If designed well, conversational user interfaces offer many benefits to end users. Language is the most natural interface humans understand. Therefore, talking to a machine rather than entering specific commands facilitates a more natural user experience. It will take a user less effort to get familiar with an application as there is no need to get used to screens, navigation hierarchies, input fields and buttons. Thanks to machine learning, applications will learn how to adapt to humans, rather than humans needing to adapt to applications.
Conversational user interfaces also save time. First, users do not need to download an application or navigate to a website to communicate with a company. Second, they can start a conversation right from a messaging app (like WhatsApp, Wechat) or digital assistant (like Siri, Alexa) – interfaces they know and understand. Third, being able to talk rather than type frees up your hands so you can multi-task. And fourth, chatbots offer a 24/7 service availability – so users can engage with companies also after office hours.
Another benefit for users is the personal touch that conversational UIs can offer. Two-way communication gives a feeling of personal attention, even if the “person” at the other end is a machine. Natural language processing and machine learning keeps improving. Digital assistants get better and better at understanding the user’s intent and context. They can remember the user’s preferences and previous interactions like purchases, complaints and requests. This information is used to generate tailor-made answers and to make proactive suggestions. And the more digital assistants are able to understand humans, the more users will trust them – or even bond with them!
But there are benefits for companies too
By introducing chatbots on popular messaging services like WhatsApp, WeChat and Facebook Messenger, companies can tap into an audience of billions of users immediately. Rather than spending effort promoting the use of an application or website, companies can easily reach their customers through existing messaging channels. For example, when KLM Royal Dutch Airlines announced to be first airline with a verified business account on Whatsapp, their CEO stated his motivation very clearly: “We want to be where our customers are and, given the 1 billion users, you have to be on WhatsApp”.
Another benefit of deploying chatbots and digital assistants lies within efficiency and cost. While it might take a couple of days or weeks to train a new employee, chatbots can be trained in hours – and they improve over time as they are fed with real-life input and feedback, thanks to artificial intelligence. Moreover, intelligent assistants can take over routine or low-level tasks so employees can focus on exceptions or high-level tasks – thus adding more value to the company.
And finally, conversational applications can be rolled out at scale, while still offering a personal, 1-to-1 experience to customers. An artificial intelligence system can talk to (nearly) an infinite amount of users simultaneously, and learn what they want. On one hand, this will allow companies to strengthen their brand widely and consistently. On the other hand, this will lead to higher customer satisfaction and more business transactions.
Designing a conversational user interface
So it’s clear that conversational user interfaces are the future of UX design. For many designers, this will be new territory. Compare it with the introduction of cinema: the first movies shown in cinemas were short and filmed with a fixed camera. With technology progressing, and cineasts starting to understand the possibilities of this new medium, cinemas became very popular. Drawing the analogy with conversational UIs: mainstream adoption might still take a few years. We probably haven’t discovered yet the full potential of this new user interface as we are still learning!
When you want to learn, and start designing your first conversational UI, here are a few paradigms to keep in mind:
- Human-like: use natural language processing and machine learning algorithms to create an assistant with a personal touch. Give your assistant personality features like gender, authenticity and humor.
- Context-aware: ensure dialog flows are logical in sequence, make proactive suggestions that match with the user’s preferences and situation.
- Intelligent: Use machine learning functionality to gain knowledge based on historic data, experience, and take action in response to new or unforeseen events.
- Multi-modal: Create one assistant across all applications. For example: starting a task on a mobile device and continuing on a desktop later, or vice versa. Allow seamless transition across platforms.
- Omnipresent: integrate with applications and platforms your users use, to lower the barrier for engagement.
SAP CoPilot is ready for take-off!
You might wonder: what are SAP’s plans in this space? Well, with the paradigm shift of artificial intelligence and machine learning, we looked at the bigger picture and designed an experience not just for the screen, but beyond the screen. Enter SAP CoPilot!
SAP CoPilot is an evolution of the SAP Fiori user experience, which is key in SAP’s UX strategy. Introduced with SAP S/4HANA Cloud 1705 in May this year, SAP CoPilot is SAP’s digital assistant for the enterprise. In its current version (1708), it is a contextual collaboration tool that sits on top of the SAP Fiori launchpad. It allows for seamless collaboration across applications while creating and sharing notes, objects, screenshots, and messages – while being aware of the business context, and letting the users collaborate within one and the same SAP Fiori experience.
With the introduction of natural language interaction in the first quarter of 2018 (refer to the SAP Fiori roadmap, page 105), SAP CoPilot will also start offering a conversational user interface. This will open up new possibilities to interact with SAP applications: talking or typing questions and commands in natural language. “Show me partially paid invoices for supplier X” or “create a sales order for customer Y” will become as common as good old TX codes like “VA01” or “FBKP”.
The SAP CoPilot roadmap foresees support for a number of natural language use cases initially, and availability will expand to more use cases and SAP applications during 2018. It will also be possible for designers and developers to create custom natural language interactions, allowing SAP customers to build their own conversational user interfaces. And thanks to SAP’s machine learning capabilities, SAP CoPilot will start to learn and recognize user behavior, and recommend best next actions.
In case you want to see SAP CoPilot in action, at this year’s SAP TechEd event in Las Vegas a demo was shown during Björn Goerke’s keynote (SAP CoPilot demo starting at 46:10). You can also watch the full 38 minutes’ replay of SAP TechEd session DX 208 – SAP CoPilot by Eugen Winschel, Product Manager of SAP CoPilot.