Peter Drucker is credited with the motto, “what gets measured, gets managed.”
Experience designers know they must attend closely to every facet of the customer's interaction with their product. Making it as easy as possible is certainly top of mind. Silicon Valley's march of progress has largely depended upon sanding away every unnecessary speed bump.
Yet, simply decreasing difficulty won’t thread the needle for optimum engagement. Progressive challenges better absorb a chooser’s attention. Finding the right balance, neither too hard nor too easy, requires more exploration and experimentation.
If only there were some way to assess and compare flow across different experiences!
Existing scales for measuring flow are not at all adapted to product testing. One frequently cited scale, in full form, requires 36 questions. This kitchen sink measures 9 factors:
[1] challenge-skill balance,
[2] action-awareness merging,
[3] clear goals,
[4] unambiguous feedback,
[5] concentration on the task at hand,
[6] sense of control,
[7] loss of self-consciousness,
[8] time transformation,
[9] intrinsically rewarding
Many of these (e.g., clear goals & unambiguous feedback) intertwine. Conceptually, clarity of goals and clarity of feedback can be teased apart. But, if the two factors co-occur more than half the time, there’s diminishing returns to asking both questions. Instead of trying to maximize length for some theory-driven agenda, industrial strength questionnaires need to be as succinct as possible. Even the shorter academic scale is 13 questions long.
In a product-market context, we really don’t want anything more than a handful — say, 3 to 5 questions. I follow a straightforward method for building a brief but reliable metric. Mechanical Turk's crowdsourced population makes it easy to explore iterations that test the published instruments’ wording, as well as variations on the same ideas. It’s possible to squeeze a measure of flow into a few questions, without intensive machine learning. A simple regression can deliver a streamlined scale.
In my studies on products and flow, I asked how well each additional flow question predicted a positive user experience. In order to do this, we need to use some metric that can stand in for a “positive experience.” After trying a few different global metrics, I was relieved to find that there was a robust relationship between flow and any of the various gestalt measures. The one question most often collected, ‘willingness to recommend” (NPS) is good enough as a workable proxy for quality of the experience. Other measures, such as “exceeds your expectations,” turn out to capture most of the same signal.