Putting Responsible Design into Practice: Part 1

Rob Girling
6 min readAug 19, 2019

We all know the tech industry must do better in anticipating and avoiding the adverse outcomes of our work. While much of the design community is committed to improvement and taking responsibility for the impact of what we create, there are few resources out there for how to improve.

In this two-part series, I will provide practical advice on how to design more responsibly. In part one, I explore what responsible design means and why it matters. In part two, I share four principles of responsible design and how you can apply them to your work.

Rewind the clock a few years to 2015. Imagine you’re working at Airbnb on the design team responsible for guests’ profile functionality. You and your team probably have no idea that what you are working on is about to become a massive problem for the company. In early 2016, Harvard professors published a paper demonstrating how the design of Airbnb profiles allowed hosts to discriminate against guests based on their photos and names. Not only did countless guests experience racial discrimination on a platform whose mission is “to create a world where anyone can belong anywhere,” but a withering two-year negative press cycle started for Airbnb, accusing the platform of not doing enough to prevent this discrimination. It’s hard to quantify the cost of all this to the company, but let’s conclude that it must have been significant. I’m confident in stating that Airbnb would have preferred to avoid this episode.

As that designer at Airbnb in 2015, could you have anticipated these challenges and changed the design? Moreover, could you have convinced your peers that the design being shipped was not ready for prime time?

Before I go any further, I must say that Airbnb handled this situation commendably after the Harvard paper surfaced. I have great respect for the company and its design team, and the ideas they eventually employed to mitigate the problem.

However, I contend that the negative consequences of many tech products and services over the past few years could have been identified ahead of time and avoided by taking a responsible design approach.

There is a chorus of people in the tech industry who refuse to accept that we could do a better job. Their argument is some variant of: it’s too hard, it would take too long, or it is not cost effective. The most annoying variant of this thinking is the attitude that it’s easier to fix something when it becomes clear that it is a “real” problem. Waiting until after you’ve done real damage to address things is an irresponsible philosophy. Many industries like finance, pharmaceuticals, and healthcare have discovered that this method quickly leads to regulatory oversight, brand damage, employee backlash, and public outcry.

It’s never been more critical for the tech industry to try and design more responsibly. It’s not only the right ethical thing to do, but it makes clear economic sense to avoid unintended brand-damaging outcomes like the above example. Let’s examine how.

Methods. Culture. Values.

Good intentions alone won’t eliminate the negative consequences of our work. We need techniques for identifying potentially negative impact, and the confidence to flag problematic designs and decisions. To be effective, we must integrate this into the design process, team culture, and values of an organization.

So what can we change to maximize positive outcomes, avoid adverse consequences, and design more responsibly?

1. Methods

What kinds of design processes and practices are teams using to identify problematic or potentially harmful outcomes during the design process?

Technology design groups have a hard time factoring complex societal behaviors like racial bias into their design output. Generally, this is a result of not having a deep enough understanding of various stakeholder interests and behaviors, however good or bad (broadening stakeholder definition is one of the four principles of responsible design that I will expand on in Part 2 of this series).

Airbnb’s philosophy is very much about establishing trust between a host and a guest. This would suggest that a principle like identity transparency should prevail. While admirable in principle, this unfortunately ignores a small percentage of hosts whose biased behavior will undermine it in practice. One of many solutions Airbnb implemented to mitigate this was more “instant-book” options, where information on identity is exchanged simultaneously with the acceptance of the booking. Another idea they implemented later was stopping hosts from accepting dates that had previously been denied to others.

2. Process

How are these forward-looking methods exercised in an Agile and Lean MVP culture where sprints and short time-frames deliberately limit the depth and thoroughness of exploration?

Most Agile and Lean methods have been positive for the technology industry, but a lack of time to understand stakeholders, plan robustly, conduct an ethical evaluation, understand systems, and explore consequences and alternatives is probably the most significant single contributor to why technology products with unintended impact are released.

Design teams today are like a newsroom under daily deadlines, with nobody working on the equivalent of larger investigative pieces that require in-depth research, interviews, editorial, and revision. We need both levels of input: the teams focused on broader, more abstract issues that lead to profound insights, interfacing with the teams in the details of delivery backlogs and shippable increments.

3. Culture + Values

How is the company’s mission showing up in day-to-day discussions during the design cycle? How might we keep a company’s mission statement and values alive as a critical dimension of design work — even in the definition of individual features?

Airbnb’s mission — “to create a world where anyone can belong anywhere” — would seem to square up to the significant issues of overcoming diversity, division, bias, and inequity. Yet it’s common that in many larger organizations, purpose and mission are much more vaguely defined and not “tested” or implemented into product or service touchpoints with consistency or rigor. A responsible technique that can help in this context is declaring specific preferable outcomes for a product, service, division, or organization. These outcomes are not wishful thinking but well-defined, long-term goals that align teams to the same vision. The result of this is to create a sense of authenticity in customer experiences, one in which brand promise and experience are tightly coupled.

So, what is responsible design?

I’ve been trying to pin down this idea of Responsible Design for a couple of years. At the highest level, I think we can all agree design is about trying to improve how things work for people. To me, responsible design has two parts: really understanding the domain you are trying to influence, and the impact of the intervention you are proposing.

In many ways, responsible design sounds like a lot like good design — design that understands its context and is thoughtful and careful about its impact. But traditionally defined, “good” design can still be irresponsible design. A beautiful, engaging and user-friendly product can be opaque in its use of customer data; exploit cognitive flaws to create patterns of addiction; ignore the risk of exploitation by bias or bad actors; and so on. Creating products and services with this kind of impact is irresponsible.

Responsible design is not a purity test, either. We shouldn’t let the notion of “responsibility” overwhelm and paralyze design with potential sources of accountability. The act of creation itself is naturally disruptive and will have effects and impact designed intentionally and unintentionally to shape behaviors, change systems, impact people and environments. These interventions can’t be fully understood ahead of time, and people’s reaction to them in the real world will never be entirely predictable. However, the onus is on us as designers to do better and approach how we think about and perform our work through a responsible design lens.

As hard as it is to create “fully” responsibly designed interventions, we must try to do better than we have done historically. Striving to accomplish more is what it means to design responsibly, a commitment do better with our work and less harm.

In part two of the series, I’ll share four principles of responsible design that you can use in your work. In the meantime, I welcome your questions and comments below.

Part Two.

📝 Read this story later in Journal.

👩‍💻 Wake up every Sunday morning to the week’s most noteworthy stories in Tech waiting in your inbox. Read the Noteworthy in Tech newsletter.

--

--