Why Design Thinking Needs a Responsibility Reboot
MIT Sloan Management Review · · g5368361
Marie Montocchio/Ikon Images For years, social media companies have come under fire for promoting divisive, false, and dangerous content in order to monetize user engagement. But in March, when a jury in Los Angeles Superior Court found Meta and Google liable for causing harm to users due to the addictive nature of their products, the […]
Marie Montocchio/Ikon Images
For years, social media companies have come under fire for promoting divisive, false, and dangerous content in order to monetize user engagement. But in March, when a jury in Los Angeles Superior Court found Meta and Google liable for causing harm to users due to the addictive nature of their products, the verdict put the focus not on content but on the companies’ deliberate design choices.
What stands out to us in considering the implications of the court decision is that even absent a willfully exploitative business ethos, other companies may unwittingly create the potential for harm simply by applying user-centered design methods such as design thinking. These methods prioritize giving users what it seems that they want. They involve engaging with target users to understand their wants and needs and having them test and offer feedback on prototype products. Our research found that such methodologies, used in good faith, can nonetheless lead practitioners to ignore potential harms or give inadequate attention to ethical considerations.
To better understand how such situations arise, we interviewed 27 senior managers and leaders at well-known technology, pharmaceutical, and design consultancy firms based in the U.S. and Europe. We asked them about the ethical challenges they have encountered when applying design thinking and other user-centered approaches, as well as the governance mechanisms they have used — or wished they had used — to mitigate unintended harm. 1
We heard about many moments of reckoning. Features that increased user engagement were described as leading to addictive user behaviors, as in the Meta case. Over time, personalization contributed to the amplification of extreme and polarizing content. In each case, the teams had followed user-centered best practices, but unintended consequences repeatedly emerged. Such experiences raise a fundamental issue: How can we better understand and manage the blind spots and ethical challenges that user-centered design may create?
Methods like design thinking have gained popularity as a creative way to identify and address user needs. 2 Applied by companies such as IBM, Meta, PepsiCo, and SAP, design thinking is a problem-solving and innovation approach founded on several principles, including collaboration across functions, iteration and experimentation, and the adoption of an empathetic, user-centered orientation.
While there are plenty of examples of design thinking as an effective methodology for developing products that appeal to users, we’ve also seen cases where its use simultaneously resulted in harm to users and other stakeholders. For example, Juul’s electronic cigarettes emerged from the work of two Stanford University product design graduate students, who used design thinking to create more appealing products than those already on the market. While the product was certainly appealing, achieving 70% of e-cigarette market share in the U.S. in the mid-2010s, it had to pay over $450 million to settle legal claims over product harm and false marketing in 2023. A big part of the problem was that efforts to improve the user experience by introducing nicotine delivery pods in fruity flavors appealed to teenage users, making Juul the dominant brand for vaping among high school students. The failure to consider their choices’ potential harms as the business scaled belongs to Juul’s leaders, but the ease with which design thinking practice can overlook such concerns must be noted.
Overall, a clear ethical framework underpinning the application of design thinking and other user-centered approaches is lacking. Our interviews suggest that this gap is systemic. It stems from recurring blind spots in how user-centered innovation is practiced.
While our interviewees acknowledged that understanding users’ needs is important, they also highlighted four issues that a strong focus on users may create.
1. Leaving the environment out of frame. Even though the natural environment was clearly affected by the design thinking implementations we considered, none of the firms we engaged with explicitly considered the environment as a stakeholder. And when someone raises such considerations, design teams aren’t necessarily responsive. The director of user experience design at a global smartphone manufacturer recalled the team’s negative reaction when, during a design session, the head of environmental policy asked whether the company really needed to release a new device every year, given how many old phones went into landfills despite the company’s recycling program.
Such issues are problematic because the creation, production, and delivery of various companies’ products and services have evident environmental impacts, including the use of rare mined materials in consumer electronics, the disposal of hazardous materials in pharmaceuticals, and the energy consumed to create digital products. In all these instances, design decisions may degrade a healthy natural environment that humans depend on. Most companies appeared to treat these issues as externalities rather than risks to be proactively managed.
2. Overlooking stakeholders who aren’t the primary target users. Concentrating on the target users often restricts product developers’ attention to specific personas, leaving out other relevant stakeholders. These may include vulnerable users, nonusers who could be indirectly affected by the product, and potential malicious users. For example, the CEO of a design services firm described a location-sharing feature in a dating app that was intended to improve matching and enhance user safety. While the design team saw it as a value-adding innovation, many women interviewed during testing compared the feature to stalking. “For features that track users, you need to design with creepers in mind,” the CEO pointed out. In this case, the team had designed for well-intentioned users, not for malicious ones or those more exposed to risk.
3. Sacrificing scrutiny for speed. Like other innovation approaches, such as agile methods and Lean Startup, design thinking entails rapid experimentation with users to develop new offerings. While quick prototyping and pressure to deliver results can be useful in certain settings, speed can reduce consideration of potential harm, especially for more vulnerable users, such as children. A senior director at a design services firm reflected on whether more time could be spent in developing and iterating prototypes with users: “In a perfect world, definitely yes. But we live and work in an imperfect world with tight project deadlines and resources.”
4. Exposing the company to risks through scaling. We found many examples of iterations conducted with a small set of users that produced dysfunctional outcomes when the products scaled and reached more diverse populations of customers, particularly vulnerable users, such as children. There are two reasons for this. First, small-scale tests are not always the best approach for unearthing a wide variety of use cases, especially malicious use cases, which often emerge as user diversity increases. Second, several executives we interviewed described how so-called edge cases were treated as statistically insignificant during development. As one expert noted, even a small percentage can become consequential at scale: “In Facebook’s case, the 1% is 1% of 2 billion users. That’s approximately 20 million people. People . Not widgets.” Features developed around a narrow user profile can behave very differently when used by millions of people with varied motivations, vulnerabilities, and social contexts.
Effective leadership and deliberate governance are required to address the limitations we’ve identified in the design thinking process. Senior managers need to understand that user-centered approaches can have ethical blind spots, and that they must take an active role in supervising their use. There are three elements in doing this.
Start with purpose, not just user needs and desires. By purpose, we mean both financial objectives and the company’s mission and values, especially as they relate to society and the environment. Potential consequences for multiple stakeholders should be discussed from the beginning, including secondary and indirectly affected stakeholders. However, there are trade-offs to be managed here. In principle, broad inclusion improves foresight, surfaces edge cases, and reduces downstream harm. In practice, it is neither feasible nor strategically prudent to convene a “committee of everyone,” as this can slow development cycles, create decision paralysis, expose confidential information, and dilute competitive advantage. The challenge, therefore, is not maximal inclusion but structured inclusion.
Three key questions should be asked at the outset:
Attend to the larger system within which design takes place. In our study, we found that processes and structures can greatly enable or limit the agency of design thinking teams. For example, in several companies, KPIs, performance targets, and rewards that prioritized time to market constrained design teams’ ability to consider diverse stakeholders and potential harms. Similarly, business models optimized for maximizing online engagement hindered design thinking teams’ ability to consider downstream harms, particularly among young people. To counter such issues, ethical considerations should be built into strategic discussions — about the business model, competitive strategy, and key objectives, for example. Decisions about user interfaces and product features must not be seen as purely operational. Leaders must promote psychological safety for designers and developers so that they feel free to speak up about potential product harms.
When considering the broad context, ask these three questions:
Pause before scaling. Before a product launch, assess the full spectrum of benefits and harms across multiple stakeholders and then decide whether to proceed. In addition to reviewing the product against its initial purpose, leaders should evaluate it against ethical criteria: Does it respect relevant rights and duties? What are its most consequential downstream effects? Does it nurture context-sensitive relationships and responsibilities?
In our study, several experts noted how difficult it is to anticipate unintended consequences before a product reaches scale. AI tools can assist by simulating edge cases, modeling malicious-use scenarios, and stress-testing assumptions. For example, a team developing a content recommendation feed could, before product launch, generate synthetic user profiles across age groups and other risk factors and stress-test whether the system disproportionately steers some users toward content that is extremist or is disinformation, for example. If it does, the team could require mitigation (such as friction, downranking, or guardrails) as a condition of launch.
Discussions of scaling risk might consider the following questions:
As products, services, and algorithms become more sophisticated and more complex at scale — reaching more people faster — the cost of getting things wrong shifts from inconvenience to measurable harm and legal liability. A more exacting approach may add time and cost upfront but could reduce downstream consequences such as reputational damage, product recalls, in-market fixes, and litigation. The question is not whether leaders can afford diligence but whether they can afford to postpone it. In some industries, those risks are no longer hypothetical — they’re already being tested in court.