Whereas once technical knowledge and rapid response were all that mattered to support users, that is now table stakes. Today, users from diverse parts of a customer’s company need to be able to access and understand support quickly and easily, changing the goals for what a support session should look like.
That is why today, SAP is using innovative functions enabled by artificial intelligence (AI) and real-time support with cutting-edge Built-In Support, built directly into the application, to provide customers with a rapid and intuitive experience that is paving the way for increasingly predictive and preventative support.
“Today, user experience matters more than ever in the support context. Users expect a seamless end-to-end customer experience, and we are committed to providing it to them by offering connected and holistic support,” said Andreas Heckmann, executive vice president and head of Customer Solution Support and Innovation at SAP. “This means meeting the customer inside the application with contextually aware features.”
This innovative customer experience leverages AI and machine learning to deliver intuitive experiences that users have come to expect from consumer applications, but with the full might of SAP’s engineering powering the interaction behind the scenes.
Here, Heckmann goes deeper on how these innovations came to be and how they can show us the future of support.
Q: There has recently been an evolution in Built-In Support as part of SAP’s Next-Generation Support approach. Could you tell us about what’s changed?
A: Built-In Support, which means support that is available within the application, has been substantially enhanced recently, including a complete re-platforming, making it a lot more feature-rich by adding AI functions and real-time chats. That means customers don’t have to leave the application to experience these cutting-edge innovations; it’s all right there for them, and it’s contextualized because we know where they are in the application. So if you ask to chat with an expert inside the application, it won’t take you to some unrelated person. The AI system will know what type of question you probably have.
How does Built-In Support impact user experience, and why is user experience (UX) or user interface (UI) so important today?
With Built-In Support, I can now interact with support the same way I interact with my smartphone as a consumer. Normally when I’m using my smartphone, I have an expectation that everything is there on the spot, easy, intuitive, fast. That’s what Built-In Support feels like to the user.
In the past, we in support mostly interacted with IT departments, which were filled with very experienced IT professionals. For them, clumsy UI wasn’t much of a problem. Nowadays, we are interacting with business users from any line of business more and more. And those users increasingly belong to an age group that consists of “digital natives,” hence having an entirely different expectation on user interface and experience. To meet all our users on eye level and provide them with the same consumer-grade experience, we invented a solution like Built-In Support, making it easy and seamless to engage with product support, no matter if you are an IT expert or an expert in your specific line of business. Our sophisticated technology is designed to back this all up.
With AI and machine learning helping to simplify, improve, and accelerate processes and workflows — just to mention a few advantages — what is unique in how SAP employs AI?
Today, already every single one of the interactions we have with customers is monitored, processed, and analyzed by our AI system because we embedded it in our incident management process. That means that customers are typing away and describing their problem and we are analyzing the likely cause in real time. They don’t even have to press a button or activate something. They just go through the normal process like they always did. That’s pretty unique because SAP is not a one product company, not a 10-product company, but a hundreds-of-products company. In that context, being able to give answers very, very precisely and to be right most of the time, that’s quite unique in the industry, and we were only able to get there by developing several of our own algorithms to work seamlessly with our proprietary systems and data.
Did your team encounter setbacks as you established this AI program?
To be honest, more things didn’t work out than did work out. When we first began using AI, we thought it would be relatively easy, just applying some algorithms. It didn’t work at all. We learned the hard way that because of the many products we have and because many different technical terms employ the same words but sometimes with different meanings, traditional approaches to AI didn’t work for us.
Take, for example, the word “transport.” What does transport mean? We have a logistics system, and within that, if a customer discusses a transport, she is probably referring to transporting items on a truck, on a ship, or on a plane. But it could also mean the transport of a piece of software from one system to another, because we have software transport systems as well. It could refer to moving data or it could refer to moving people.
That was the nut we had to crack. And we had to learn that AI is always made for the data you are using. At the very beginning, I didn’t get AI. I looked at it more like a search engine, a better search engine with a smarter filter, but that’s completely wrong. For any type of data and every use case you must build your algorithm and then you must train your algorithm. You must find the patterns within. Each situation is somewhat unique. That’s the biggest lesson we learned.
How have AI and other emerging technologies paved the way for predictive support?
AI is a bit like a key ingredient that we need to prepare the meal, which is predictive support. All the things we’ve been discussing — real-time support, AI, and so on — all of it is the innovation that brought us to the current status quo. So why do we already regard this as the old world? Because the customer is still experiencing the problem, and we are just getting much better, faster, and more creative in how we respond.
What we are trying to do now and for the time to come is combine real-time support and AI and other innovations to find the problem before the customer even knows they have it. Say a customer reports a problem, and then another customer reports that same problem, a short time after, and maybe a third customer reports the problem. Today, we will automatically detect that and conclude, “Hey, something’s happening. We have three customers reporting this type of problem. What can we learn?” And we look automatically at these customers and discover they are all using a specific functionality in a specific configuration. And in that constellation, they are running into that problem. Then we can automatically determine which other customers have that module, use it in that very specific constellation and solve the problem for them before it even occurs, and provide a predictive and preventative service to help those customers before they even experience the problem.
