Join Transform 2021 for the most important themes in enterprise AI & Data. Learn more.
Ocean Spray Cranberry company has always used a human touch when interpreting data about its customers. The company developed its wildly popular Craisins product after discovering that only 38 percent of U.S. consumers were eating or drinking cranberries but around 60 percent of Americans like dried fruit.
When social media became popular, Michael Nestrud, the company’s senior manager of global sensory science and consumer affairs, made a hobby of analyzing Twitter posts about cranberries to get a sense of customers’ emotions associated with the fruit.
In many ways, Michael’s approach is the wave of the future. However, unstructured customer data is unstructured and anecdotal — only useful when you are able to sense patterns and incorporate the insights gained from them into customer outreach and experience.
Ocean Spray recently adopted an artificial intelligence (AI) platform that senses patterns in customer behavior to sense market conditions and then enable adjustments to price and targeted promotions designed to delight the company’s fans. While impressive and incredibly powerful, such AI platforms do have one fundamental flaw: They’re machines, not humans.
Integrating design thinking
Ocean Spray’s data scientists need to make sure its predictions sync with employees’ understanding of what makes customers tick. They’ll also need to think carefully about how the AI system’s predictions and recommendations will be experienced by their customers, whose emotions and emotional associations the machine can’t effectively interpret or predict. Heading into the new year, companies adopting AI platforms have a real opportunity to incorporate those human traits into making decisions. They’ll need to embrace design thinking.
AI systems like these are unable to tap into empathy for the humans they’re analyzing, but integrating design thinking into the process infuses empathy and humanity into the system. Doing so is often a matter of situation-specific problem-solving.
Imagine a scenario where a random fad for cherry relish reduces customers’ interest in cranberries at Thanksgiving one year. An AI system may recognize that customers who usually share cranberry recipes on social media around the holiday are sharing less of them than usual, or that they are using the word “cherry” alongside the word “cranberry” in a surprisingly large number of posts.
The AI system may recommend offering a discount on cranberries in certain regions where the fad is strong and cranberry-related chatter is low. But with an understanding of humans’ interest in novelty and the emotional appeal of tradition, design thinking might inspire an ad campaign revolving around how to make cranberry-cherry relish or emphasizing family members’ probable disappointment on Thanksgiving if their favorite traditional cranberry sauce isn’t served. It might even recognize that we, the authors, love cranberries — because, admittedly, we do!
This human-level interpretation makes design thinking tough to duplicate at scale. However, AI is designed to learn these patterns and trends. With each iteration of human intervention into the system, it can begin to incorporate new connections. Next time another similar fruit starts showing up next to “cranberry” in a lot of Twitter posts, the system may automatically recommend taking action that inspires customers to either combine the two fruits or to choose cranberries instead on Thanksgiving — without Ocean Spray having to lower the price.
The power of computer-human partnership
In this collaborative approach of design thinking and AI, each approach contributes its strengths and minimizes its weakness to deliver a more efficient development process — and a better user experience. AI systems gain a level of empathy and predictability — such as an understanding of why those cranberry searches are happening versus just that they are happening — while design-thinking practitioners get help with heavy lifting tied to data analysis, predictive analysis, and problem-solving. In this case, they’d be able to see and interpret the social chatter without having to wade through endless posts and hashtags.
Once established, these systems can improve and enhance over time, gaining from the progress and insights collected during the development process. It’s one of the unique ideas powering AI-driven design thinking — AI has to constantly learn in order to get better, while design thinking leans on audiences to gain consumer insights.
Together, a combined AI/design thinking approach can lead to the creation of products, services, and niches that most accurately align with customer wants and needs. What’s more, this process can get to the desired end goal even if customers can’t articulate what they want — and it can get there without endless customer research and trial and error.
David Parmenter is Director of Data & Engineering at Adobe.
Liang-Cheng Lin is Sr. Experience Design Manager at Adobe.
VentureBeatVentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact. Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:
- up-to-date information on the subjects of interest to you
- our newsletters
- gated thought-leader content and discounted access to our prized events, such as Transform
- networking features, and more