A look at research showing how accountability shapes outcomes, why those outcomes aren’t as good as we think, and what we can do about it.
If you work in a design-oriented role, to what extent do you agree with the following statement, as it relates to you personally?
I approach my design work so that it is made to be in fact effective, rather than to look effective.
If you’re like me, your response is pretty much automatic: Strongly agree.
But I want to challenge us both on how often that is actually the case. To do that, we’ll examine research on how accountability works, who designers are accountable to, and how it all makes the above statement harder to live up to than it might seem. Then I’ll end with some ways that might make things easier.
How accountability influences thought
First let’s lock in what is meant by “accountability.” Philip Tetlock and Jennifer Lerner define it as an “explicit expectation that one will be called upon to justify one’s beliefs, feelings, or actions to others. Additionally, accountability implies that positive or negative consequences hinge on the acceptability of one’s justification.” Tetlock and Lerner published research on accountability showing us that when people know in advance that they’ll have to explain their decisions to other people, they become much more diligent. They don’t rely on intuition alone. They take a reasoned approach. They do research, include evidence, make revisions in accordance with that evidence, consider alternative approaches, and think in terms of systems— key concepts in what we call “design thinking.” All this stuff increases what psychologists call exploratory thought, reasoning that “considers multiple points of view and tries to anticipate all possible objections to, or flaws in, a particular position.” Accountability is a kind of cure for laziness.
So when designers are accountable, they engage in exploratory thought, and exploratory thought should lead to good design decisions and outcomes. With that in mind, let’s look at the accountability forces on designers.
Who are designers accountable to?
Anyone operating in a design-oriented capacity — including visual, UX, and product designers—are accountable to someone. I’ll argue that this “someone” is actually a collection of humans. Some of the relationships in this collective are direct, some are more of an indirect dotted line, and others have a kind of backdrop presence.
These are the people with the keys. Things don’t ship if they explicitly withhold approval.
- Stakeholders have strategic or organizational goals to achieve.
- Product owners/managers have their name on the product, executing on a high level roadmap.
They can’t really block your work, but their buy-in has high political value.
- Project managers are working to see the product design done within resource, time, and budget scope.
- Managers might peer over your shoulder, or provide guidance, but shouldn’t actively filter or mutate your work (let’s hope).
- Influencers might be found within or around your team whose good favor or opinions are sought and valued.
- Developers and engineers need to understand, believe in, and implement your design.
These people are omnipresent. They may not have visibility of your work. But if they do, you care what they think of it.
- Peers are your colleagues within or near your team.
- Mentors might be outside your organization.
- The design guild is the community of practitioners within and around your craft, all around the world, to whom your work is visible in some way.
This is your field of accountability. You can feel pressure from different points in this field, at different intensities, and at different moments in time. Because they create a demand for reasoned explanations, it’s sensible to think that these accountability forces make for conditions that yield better design outcomes.
And that’s true. Except when it isn’t.
When accountability backfires
It might be easier to discuss when it doesn’t backfire, because it turns out there’s a very important twist in Tetlock and Lerner’s research: Most of the time, accountability results in the opposite of exploratory thought.
Tetlock and Lerner showed that exploratory thought only happens under certain conditions: When the person accountable…
- Knows in advance that they’ll have to explain their thinking to an audience.
- Does not know the preferences of that audience.
- Believes the audience is informed and desires objectivity and truth.
When those conditions aren’t met, what happens instead is confirmatory thought, or the kind of thinking where a desired outcome is already targeted, and the reasoning happens afterward in service of that outcome.
In other words, if you know you’ll have to explain yourself, and you know what your audience wants to hear, your brain wants to use its reasoning not to find the thing that is right, but to justify the thing that sounds right. Confirmatory (or motivated) reasoning, then, is really about being persuasive, about looking good over being good. In fact, confirmatory thought is so convincing that it works on not just on your target audience, but even on yourself. Your own mind can become convinced by its own confirmatory reasoning. Politicians do this all the time.
Designers expect to defend their decisions. However, the other two conditions for exploratory thought usually fail. Audiences have preferences that are known to a designer, and are likely to support design that maps to their own desires over design that serves objectivity. Because design happens in conditions that almost never comply with those that allow exploratory thought, what often gets triggered instead is confirmatory design, or what I call motivated design — design not through reasoning informing the decisions, but exactly the other way around. Motivated design can be as bad at producing good products as motivated reasoning is at finding truth.
With this in mind, recall all the forces of accountability acting on design. The stakeholders, the guild, peers, influencers, and the rest. Think of their preferences and desires. What do they tend to value? What are their incentives? What kind of design are they likely to reward?
It’s important to remember that everyone behaves differently as an individual, and almost always with good intent. But here are a few reward tendencies associated with a group:
Design guild: Design that has the look of the moment, looks good on Dribbble, or optimized out of context.
Developers: Design prioritizing minimized effort or repurposing pre-existing work.
Managers: Design that’s unthreatening to the role or ego of smart advisor.
Project managers: Design that’s inexpensively produced.
Product owners: Design that accommodates any ask (or inversely, tunnel vision).
Stakeholders: Design that appeals to personal preferences. Design that achieves a goal at the expense of long term viability.
These forces can shape design, founded on persuasive justification, without anyone (even you) realizing it’s happening. What results is work that is a subset of what would have otherwise been explored, optimized for a biased field of accountability. Work that is designed to be rewarded by a certain audience. Work that probably looks good, but might not in fact be good.
Exploratory thinking in a landscape of accountability
Truly operating in realm of exploratory thought is hard, but I think it can be done. To help, here are some ways of thinking and questions to ponder.
- Come to terms with the human mind’s relationship with accountability. Just knowing the conditions in which confirmatory thought happens can help detect it before it leads you too far down an unsubstantiated direction.
- Envision a design as a product being sold, where the only buyer is the end user. You don’t have to convince a shopkeeper in the middle to buy it beforehand. Would that change the design? How?
- Routinely stop during pursuit of a design direction and ask: Which audience in your field of accountability would offer the highest bid for this direction, and why would they do that?
These questions are aimed at identifying the real reasons driving a direction, and to recenter those reasons on objective reason and ultimately the user of the product.