By John Campbell / 08.18
The Uncertainty Principle
Poetry meets product development.
For better or worse, I’ve fallen in love with the poet John Keats’s notion of negative capability—that rare and strange ability to comfortably exist in a state of uncertainty. Or as Aileen Ward, author of John Keats: The Making of a Poet put it, the capacity to “remain content with half-knowledge.”
In true Keatsian fashion, the term was nearly a throwaway. It only appears once in his letters:
But Keats’s pithy aside still resonates, and not just among poets. Social theorists, philosophers, and psychoanalysts have all re-interpreted “negative capability” to describe different skills and states of mind.
As a content strategist, I’m also trying to render Keats’s idea my own. I work with designers, product managers, engineers, researchers, and others to build and refine parts of Airbnb that are central to the guest experience. I spend a lot of time searching for and testing elusive content frameworks and text strings that’ll make the numbers sing. But oftentimes, at the end of an experiment, the reasons why a particular variant performed best remain fuzzy and equivocal. Distilling broad truths from the results can feel like educated guesswork.
So, the reluctant poet in me finds the value of negative capability clear—it validates the grey areas of my work and makes it okay to occupy the space between concrete results. But what about product managers and data scientists? Could a sense of negative capability also help my co-workers whose careers depend on something closer to philosophical certainty?
The persistence of doubt
In the tech world, virtually every product launch is informed by some kind of statistical assessment or hypothesis test. Product teams experiment with different versions of a feature, look for changes in certain metrics, and, using statistical analysis, determine and launch the version that performs best.
Statistics might seem antithetical to negative capability. But as any data scientist will remind you, complete confidence in any result isn’t feasible, or even desirable. At a 100% confidence interval, most experiments wouldn’t be possible, or would last indefinitely. So, however small, experiments—and therefore product development—depend on some measure of doubt.
So long as teams stay curious, uncertainty isn’t cause for alarm. What we don’t know leaves the door open for new and better products.
Acknowledging this doubt and its limitations are table stakes for every statistician. In fact, most analysts and data scientists I know are quick to point out the deficiencies of an experimental design, or their ability to draw conclusions from it. So, in a way, my teammates most explicitly tasked with mitigating doubt are already quite comfortable expressing that doubt. And they have a bespoke vocabulary for it, too, with concepts like statistical significance, confidence intervals, and p-values at their disposal.
A keen sense of negative capability, then, might be a prerequisite for every great product team. Which is also, if we follow Keats’s prescription, the disposition of every great artist.
Uncertainty as inspiration
Among poets, “inspiration” is a suspect word—it’s over-attributed, aggrandizing, and can diminish the role of hard work. Uncertainty and doubt, however, feel like old friends. The irony here is that feelings of uncertainty and doubt are fonts of inspiration for many poets. So, if we squint hard enough, uncertainty and inspiration can start to look like the same thing—synonyms, even, in some poetic way.
A way forward
If this article is attempting to reframe negative capability as a tenet of product development, then it needs a secondary meaning: the capacity to see uncertainty and inspiration as two sides of the same coin.
For some teams, negative capability may simply be a new term for how they already operate. For other teams, embracing negative capability could change how they view and act on experiment results, engage with research, or structure team goals.
Whatever your persuasion—content strategist, data scientist, researcher, product manager—your success hinges on your ability to live with uncertainty and yet act decisively. To offer a way forward. That way forward just might reside within your sliver of doubt.