the density of space

I’ve been playing around with Matt Webb’s motivations flowchart, a useful inversion of the typical way we design processes. We begin by modeling users as a state machine of motivations and goals (using research to validate our assumptions) and then distill a process from the model. We look for the seams of the process by identifying potential halting states in the user's motivations and exploring ways to bend users back into the system once a halting state has been triggered. By diagramming user motivations as state machines, we can evaluate how best to fulfill user needs by smoothing the path between motivations, improving the stickiness of our application or process as a whole.

Recently, my wife and I have started getting into organic gardening, and we’ve come across people trading and sharing seeds. The Organic Gardening forum has a highly evolved system for facilitating seed sharing—they have designated seed “banks,” individuals (geographically distributed by growing region) who aggregate seeds to share with others by request with the expectation that those who “borrow” seeds will eventually contribute back to the bank as they are able. Initial contact is mediated through the forum but eventually switches to email. Banks post their inventories to the forum, and individuals post their seed requests.

Assuming users are interested in one of two tasks, finding or sharing seed, I see two issues with using a forum to mediate these exchanges. First, it is difficult for a newcomer to identify the “banks” and review their current inventories without some digging. Second, it is more difficult for a “bank” to maintain a current inventory (as they have to post a new list every time there is a change). Both of these issues introduce substantial drag on finding and sharing seed. In practice, the community around the Organic Gardening forum is motivated enough that these issues don’t (observably) prevent seed exchanges. Moreover, by focusing exclusively on finding and sharing seed, I’m ignoring the potential for unplanned community interaction inherent in a forum. I’m not interested in replacing the existing system—I just want to take Matt Webb's technique out for a spin on a non-trivial problem.

Below, you’ll find two of the flowcharts I constructed—Wants to receive seed and Receives seed from someone. The start and end states are in bold, and states that are independent of the particular flow are in grey. In constructing these flows, I thought of the end states as attractors, pulling the user through the process. The end states have weight, so they exert a downward pull. In designing these processes, we want to create a smooth gradient towards a desirable end state. To accomplish this, we must remove any drag or barriers that would discourage or prevent a user from flowing towards the desired end state.

Wants to receive seed flow

Receives seed from someone flow

The flowcharts above illustrate two things—first, while registration occurs only once and login occurs at most once per session, the user needs to smoothly (and transparently) transition through the registration and login states to ensure that they do not get bogged down and abandon the process before reaching their goal (this observation should be intuitive to UX practitioners). Second, for this process to be sticky (i.e., to encourage users to return), we need to explore ways to transition from the end states back into other flows (e.g., going from Contacts seed owner to Receives seed from someone) or ways to convert one end state into another more desirable end state (e.g., going from Idle to Contacts seed owner by notifying a user that someone has just posted a seed to their inventory that the user wants). We could convert these seams into business opportunities as well (e.g., offering users the opportunity to buy seeds they can’t find through a retail partner).