For Radical Compassion
If empathy is not enough to ensure responsible product & service design, we need a new paradigm. What is it, and how do we get there?
Human-centered design operates in an empathy economy, where “empathy” is a primary resource. We empathise first, to ground innovation in real problems experienced by real people, and those problems become opportunities for commercial and non-commercial service and product development to make life better.
Even though many of us (including me) have adopted and promoted the argument that empathy is essential for good product and service design and delivery, empathy alone is not enough to ensure responsible, sustainable, or equitable new product/service development.
As Hareem Mannan recently wrote in this excellent piece, there is “inherent and undeniable power and privilege in being able to shape [designed] experiences for the rest of the world.” We must not use “empathy” as a crutch to explain away our privilege, to justify our power, or to effectively erase differences between diverse humans and experiences.
In the HCD economy, empathy is commodified; it is treated as a resource that can be mined and refined by experts. When empathy is treated as a deliverable commodity, it can be consumed by people who have little to no direct exposure to the people whose problems they’re solving. We produce artefacts that mediate between users and our stakeholders. Rather than creating empathy for users, these artefacts can sometimes reinforce assumptions and stereotypes, especially when they’re poorly constructed or built on limited data.
Even when they’re well researched and designed, empathy artefacts are mere representations, and letting representations stand in for people can be extremely dangerous. It is entirely possible to use our understanding of people — captured and reflected in highly crafted artefacts that help our stakeholders empathise with users — to get users to behave in ways that benefit us rather than them. Hooks, nudges, and dark patterns are a few ways that our understanding of human cognitive tendencies and shortcuts can be deployed against the better interests of product and service users. How well do the artefacts we use to speak for “users,” speak up and push back against exploitation?
So, if empathy and its deliverable artefacts are potentially misleading or dangerous tools, what can we do to put some constraints around their misuse?
A couple of recent books argue that empathy is limited and/or dangerous, and that we should replace or go beyond it with “compassion.” There’s something really intriguing about compassion as an addition or alternative to commodified empathy, and I’m going to engage in some word play to come at this from a different angle. In short, we need a radical compassion that helps to collapse the distinction between us (the designers and the organisation) and them (the users of the product or service we provide) if we’re going to get closer to a just and sustainable approach to service/product design and delivery.
Before I get there, let’s take a tour through two seemingly very different arguments about how and why compassion should replace or extend empathy: Paul Bloom’s Against Empathy, and Eric Meyer and Sarah Wachter-Boettcher’s Design for Real Life.
Against empathy: rational compassion
Bloom’s work (I haven’t read his book yet — I’m working from a published interview and some reviews), seems to argue that empathy is flawed because it is too reliant on feeling, in addition to leaving us susceptible to bias and tribalism. For Bloom, compassion is preferable because it is dispassionate and compatible with more rational decision making.
A very long quote on this, from a recently published conversation between Bloom and Sean Illing of Vox:
“By empathy I mean feeling the feelings of other people. So if you’re in pain and I feel your pain — I am feeling empathy toward you. If you’re being anxious, I pick up your anxiety. If you’re sad and I pick up your sadness, I’m being empathetic. And that’s different from compassion. Compassion means I give your concern weight, I value it. I care about you, but I don’t necessarily pick up your feelings.
A lot of people think this is merely a verbal distinction, that it doesn’t matter that much. But actually there’s a lot of evidence in my book that empathy and compassion activate different parts of the brain. But more importantly, they have different consequences. If I have empathy toward you, it will be painful if you’re suffering. It will be exhausting. It will lead me to avoid you and avoid helping. But if I feel compassion for you, I’ll be invigorated. I’ll be happy and I’ll try to make your life better.”
So, one risk of empathy is that it can be overwhelming to feel the pain of other people and may lead to avoidance. Compassion, on the other hand, induces less pain in the observer, so helps them maintain the energy to help.
I’m not particularly sold on the idea that “compassion” is really that functionally distinct from “empathy,” despite the appeal to neuropsychology. It feels like a bit of clever word play to create an opening for a specific argument. I should probably read the book. (Incidentally, Indi Young has a transcript of a podcast in which she was interviewed critiquing Bloom’s Against Empathy — it’s worth a read!)
Reading from a review by Jesse Singal at The Cut, Bloom’s focus on the relative emotional weight of empathy versus compassion does seem to be about building an argument for rationality. Singal reports Bloom arguing that, while “compassion” maintains a level of emotion, it is far less than “empathy,” and that Bloom’s argument is about combining “the humanity of compassion” with “the rationality of cost-benefit analysis” into a “rational compassion.” Compassion here is defined as “‘simply caring for people, wanting them to thrive’ … a general feeling of goodwill toward our fellow humans.”
I’m also not sold that a “general feeling of goodwill” is enough to keep us accountable, and an appeal to “rationality” via “cost-benefit analysis” leaves me worried about how we’ll measure the costs and benefits, and what those metrics will lead us to overlook or ignore. Utilitarianism, if that’s what this is, can be incredibly narrowly focused.
Beyond empathy: practical compassion
Moving in another direction, Design for Real Life presents a compelling set of examples of what happens when organisations forget (or ignore) compassion. Here, compassion is presented — quoting a talk by Karen McGrane — as a “much deeper level” of empathy, in which you “have genuine emotional feeling for the struggles someone is going through and you are spontaneously moved to help them because you feel them” (p72).
Rephrased, the authors write that “compassion is more than just being nice. It’s accepting people as they come — in all their pain, with all their challenges — and not just feeling empathy toward them, but doing something with that empathy. It’s recognising that users facing stress and crisis need more than our sympathy. They need our help.” (p72).
So, contrary to Bloom’s argument that empathy involves too much feeling, Meyer and Wachter-Boettcher see compassion as more emotionally engaged than empathy. Compassion here also seems to involve taking action, driven by the deeper emotional resonance of compassion over empathy. Compassion is empathy, but deeper and put into practice.
In addition to compelling case studies of compassion fails (and efforts to recover from those fails), Meyer and Wachter-Boettcher give readers a rich, accessible playbook for incorporating compassion into our product research, design, and delivery work. The second half of the book consists of a series of chapters that cover how (and why) to cultivate compassion, learn from users, humanise the design process, and build support for all this within organisations that might not be ready for it.
Underlying all of this is a foundation of collaboration within the organisation. To quote from the chapter on humanising an organisation’s process (page 104):
“One thing you may have noticed about each of these techniques is that they’re fundamentally cross-discipline: design teams talking and critiquing one another’s work through the lens of compassion; content strategists and writers working with designers and developers to build better forms and interactions. Wherever we turn, we find that the best solutions come from situations where working together isn’t just encouraged, but is actively built into a team’s structure.”
Meyer and Wachter-Boettcher provide valuable advice for introducing compassion into interactions with humans in the form of better, more sensitive research and deliverables. They advocate for increasing the fidelity of user artefacts by including human imperfections and recognising that we’re not always at our best when we use a product or service, calling on designers and organisations to plan for stress cases. They also document a number of excellent techniques for focusing on real users and real situations in critiquing and testing design and delivery work, all of which improve on some of the compassion-challenged design and delivery practices that exist in large and small companies.
Us… and them
Using compassion to improve collaboration inside an organisation is an important avenue to pursue. It’s clear that many organisations struggle with internal collaboration — sometimes feeling an awful lot like a Game of Thrones in practice — and overcoming silos and politics is itself a massive undertaking. But achieving unity (and even increasing diversity) inside an organisation may still leave intended users and beneficiaries on the outside, generally unable to look in at what is being developed, how, and why.
Despite the shift from empathy to compassion, I’m still concerned that a focus on better artefacts leaves the door open for exploitation.
There’s a subtle but important issue of power here: organisations — and designers inside organisations — solve problems for people, mainly in order to generate revenue and create profit. If we continue treating people (with as much compassion as possible) as sources of problems, stimulus for design thinking, and limited sounding boards to test for problem-solution or product-market fit, we risk taking for granted that our solutions will actually benefit them and will be perceived to do so, and that whatever commodity we demand in exchange for service (whether it’s money, data, or time) is justified.
There’s no inherent protection against using the deeper, more genuine emotional connections with problem-havers to shape their behaviour in ways that tip the benefits far to the solution-producer side of the equation. It is certainly possible to mis-use compassionately extracted information to produce human-like artefacts that help a company simultaneously support and exploit human weaknesses (not least among them, those data-harvesting platforms provided under the guise of “free” services).
And again we find ourselves with distance between an us (the organisation and/or designer) and a them (the people it is ostensibly working to support). Bloom’s argument looks to be about creating and maintaining safe emotional, psychological, rational distance. More separation, not less.
Meyer and Wachter-Boettcher appear to be comfortable with maintaining a smaller temporal and physical distance between designers/developers and product/service users. Direct interaction happens in controlled environments (research and testing situations), and the results get synthesised into representations of real people, which in turn stand in for those people when a team needs to think like or make decisions that affect those people.
Where Bloom is reportedly advocating for “rational compassion,” I’m proposing a more radical compassion. To get there, let’s start with a bit of word play.
Empathy, sympathy, and compassion all derive from a common Greek root pathos (suffering, emotion, feeling), with compassion having followed a path via Latin into modern English. We use “compassion” to talk figuratively about feeling sorrow due to the suffering experienced by another person. Figuratively because the literal interpretation of the word’s Latin ancestor com (with, together) + pati (to suffer) is to “suffer with” rather than just feel bad about someone’s suffering.
What would happen if we extended that literal interpretation of compassion and made it about “suffering with” people in practice?
To really “suffer with” the people we hope will benefit from our products and services, we’ll need to bring them much closer than we do today. We will have to rely less on mediating artefacts that represent the people whose problems we’re hoping to solve, and more on direct interaction. In its simplest form, this might mean much more up-front problem-space exploration and much more concept testing and iteration with intended users.
Interestingly, this kind of direct interaction — radical compassion in the form of suffering together — is at the core of the successful, cross-disciplinary collaboration that Meyer and Wachter-Boettcher prescribe for transcending silos and creating great, more human (and humane?) organisations. By working together more, we learn to understand, appreciate, and value the contributions of our colleagues and cross-disciplinary counterparts. Imagine what we might achieve if we extend direct collaboration further, to other “Others” in our product/service ecosystems?
But is increasing the volume of interaction enough to ensure we make ethically sound decisions? Or do we need to go further? Could we bring more people into interactions more often and still treat them as exploitable end users?
Hypothetically, an organisation like Facebook could have spent extensive time with users exploring and testing the Newsfeed to understand whether it was worth releasing into the wild. In those interactions, what would it have taken for the company and its representatives to say: “we know that the Newsfeed uses the principle of variable reward to keep you scrolling and that you’re likely to feel less in control of your behaviour and time; we think that’s worthwhile, because when you stay in the Newsfeed longer, we can record your interactions and show you more ads, which we’ll use to generate a huge amount of revenue, expand our platform to try to swallow up all of the internet, and buy up rival companies whose products you might choose instead.”
And if Facebook had been so transparent about its business-model and product mechanisms, what might the response have been from users? And what might Facebook’s response to that response have been?
Like with the mediating artefacts that represent “users” in design and business decisions, focusing on increasing the amount of carefully controlled interaction between organisational insiders and intended users may be less than we need to ensure more sustainable, just, and equitable relationships and outcomes. We may need to include radical transparency as a practical component of radical compassion.
Barriers to radical compassion?
There are some obvious and important barriers to implementing radical compassion today. For one, it will increase the cost of research and testing (more of it, with more people) — and given that in many organisations it is still a struggle to get even adequate research and testing done for new (and existing) product and services, that alone may keep researchers and designers away from this approach. As valuable as research and design (thanks McKinsey…) are for organisations today, it is much less expensive to run small projects and produce artefacts that simplify and stand in for users/customers than it is to continuously and extensively engage with them in meaningful ways. Better the artefact you can deliver (under budget) than the relationships you’ll never get budget to develop and maintain.
There are conceptual barriers to radical compassion as well. As long as we continue to build and enforce categories like “users,” “customers,” etc, we are likely to be stuck in the mental model of conducting relationships across the boundary of the organisation according to exchange principles. Exchange (of money, data, time, attention or other measurable commodities) mediates the relationship between an organisation and its user/customers, just as artefacts mediate between human-centred designers/developers and the humans we ostensibly centre.
Radical compassion needs radical transformation?
In the logic of late capitalism (and in the neoliberal utopia) “the market decides” whether a product or service is viable. But we know that loopholes and inequalities across the market lead to unsustainable and unjust outcomes.
If radical compassion, including radical transparency, is going to be a viable solution for the problem of sometimes (often?) exploitative application of “human-centred” design, we may need to completely reconsider the structural relationships among the parts of a product or service ecosystem.
True collaboration will likely require a level of openness and access that is entirely uncommon (perhaps completely absent) in the world today. Would it be enough to form user/customer groups empowered to vet (and potentially reject) new products or services on grounds of sustainability, justice, or equity concerns? Could that sort of power be successfully delegated to government bodies or other organisations? Or would we effectively need to dissolve the boundaries or organisations and their constituent communities? Could or should anyone (or anything) whose life is impacted by the production and delivery of a product or service have a stake in the decisions about whether to produce it at all? The complexity of this is mind-boggling — as The Anatomy of an AI System illustrates, for a start.
If we want to really open data and processes around product/service development and organisation operation to collaboration beyond the organisation, we may need to collapse the us-them binary that is so ubiquitous today. Even if categories of “producer” and “consumer” remain technically accurate to describe roles in a transformed ecosystem, radical compassion may lead us to recognise that there is no producer without the consumer and no justification for a “producer” creating and exploiting “consumer” dependence on a product or service for private gain.
And more questions…
Beyond just the complete transformation of late capitalist political economy, we’ll have more questions to answer. Among them: how do we practice research, design, and development in a radically compassionate way? What, if anything, becomes of the artefacts we know and love as shortcuts to design thinking? What else will need to change to bring this all about? And what unintended consequences might these radical transformations have?
And more urgently, where are the examples of these transformations already underway?
I’ve got a bunch of questions about the limitations and unintended consequences of treating empathy as a commodity, and I want to explore these further with by learning from other designers, researchers, product leaders, developers, and other people about their experiences with this. (So get in touch on Twitter — my DMs are open.)