Design Thinking – Dashing from a Vague Idea to the First Prototype
Or: The art of just being creative
How does ingenuity work? How are unique products invented? By a stroke of genius, an idea that suddenly comes to you in the middle of the night as legend has it the ring-shaped benzene molecule popped into August Kekulé’s dreams? How about a toga-draped Muse whose kiss of creativity can lead a curious mind out of a dead alley of inquiry to discovery.
Truth be told, you don’t have to rely on ancient Greek goddesses to find the path to enlightenment. There are several well-known methods to boost creativity such as brainstorming or drawing up a mind map.
They fall short, however, if the challenge lies in finding solutions for complex problems in as little time as possible. That’s when approaches and concepts such as design thinking, service design, lean startup, business model canvas and value proposition design are clearly more effective. They work as innovation boosters and can turn a flash of inspiration into a veritable intellectual thunderstorm.
Creativity, design and thinking
Deutsche Telekom, too, relies on techniques like design thinking. Applying this type of mindset helped to develop two prototypes of a digital assistant during a four-day workshop, including real-life testing at the end of the design sprint.
A design sprint needs a good team made up of capable members with diverse skills. That’s why the participants at the outset agreed on a set of rules and shared understanding what they wanted to build. Key questions to flesh out in order to be successful were what they could and what they wanted, to accomplish.
The workshop’s goal was to develop a digital assistant that would improve the user experience in the online store with better usability and targeted customer interaction. But what would such an assistant look like, and what should its capabilities be? The best way to get answers is to look at questions like these through the customer’s eyes. That’s easier said than done, though, since the sprint participants all work for the vendor and are therefore seated at the other side of the table.
Up next was tapping into expert knowledge about the sales process to find out what exactly happens in a real-world setting. The answer is the acronym AAOC, short for activate, analyze, offer and close. Customers come into an online store with a certain set of needs and motivations. A vendor has to build trust and present products and services that fit those needs. If he’s convincing, the customer will sign a contract or place an order for a product.
If a digital assistant can translate this sequence of events into the setting of the online store, he (or she) should be able to proactively approach a customer to build trust. It’s the precondition to analyze customer needs and come up with relevant offers. Goods and services must be tangible, or else visitors won’t close the deal and hit the buy button.
Plugging AI into the sales process
There were more questions to consider. Does a customer even perceive the need to get advice? What exactly do customers want? And how much time are they willing to invest in this process? What’s more, participants had to clarify from the vendor perspective what factors made a difference in closing the deal: how do you define success, which interfaces work and which don’t? What does trust mean in digital sales? The final question almost borders on heresy: Do we need AI in all of this — what’s the value add for customers and Deutsche Telekom?
Since it’s impossible to answer all of those questions in a few days, workshop participants grouped them into clusters and then voted to identify key themes to turn into concepts. Their guiding principle was an idealized world in which their concepts had to function or fail. The thought exercise yielded eleven different concepts — still too many to turn into prototypes and test. That’s why the workshop team discussed the pros and cons for each idea and decided as a group which key aspects should be integrated into the first prototypes.
Eventually two prototypes materialized — rough but tangible outlines of two actual digital assistants. It wasn’t so much about developing a perfect AI but experimenting and starting a conversation with actual users. Analyzing those tests will shed a lot of light on many assumptions, confirming some and discarding others, while uncovering the potential for each concept.
What the customer wants
At the end of the design sprint, six testers met the two new AI assistants. They had been selected based on personas, but even though they represented different target audiences they behaved very differently. While one group showed more unease interacting with the “unknown power” of AI, others didn’t have any second thoughts at all.
One of the testers, let’s call him Joe, chatted with the assistant about his life and his family. He immediately showed his Fire Phone and then told the AI that he had an iPhone at home — “but only an iPhone 5.” He said he wanted to build long-term business relationships and expected fair offers. In order to consider switching, his mobile number must be easily ported and, most importantly of all, service would have to be personable.
The testers’ feedback about the assistant was as diverse as their interactions. They confirmed some assumptions of the team while shooting down others and in the process helped them evaluate the various concepts. Overall, the process showed how valuable this toolset is to rapidly develop ideas and toss them out just as fast, without wasting lots of time and money to reach that decision.
Insights matter more than concepts
The workshop wrapped up by participants presenting their results to the rest of the team. What worked and what didn’t? Where was a good starting point for more iteration to improve and fine-tune a concept?
Who’s completed a design sprint will praise it as an intense experience to approach questions with a fresh perspective. The setting is “high energy". Participants, in short, constantly motivate and push each other. The common verdict after four days: It doesn't feel like work because it lets everybody step outside their daily obligations and routines and just be creative.
Design thinking, after all, is less less about latching onto concepts, but to gain valuable insights. Those thought experiments often create more questions than they answer, but they yield a crucial result. Participants develop a shared understanding what they’ll want to work on in the months to come and they get to know their customers better.