How do you streamline a food order workflow for a mobile device?
As I studied the feedback from our end users, I found that when we design for small mobile devices, such as phones, we needed to remove all unnecessary information so that the server could convert their actions to muscle memory. Designs needed to be clear and uncluttered. We also needed be sure that tap targets were sufficiently large, to help users avoid unintended selections. I also learned the following takeaways:
1. Isolate which tasks occur once per table, and which interactions occur multiple times per table.
2. Give the user a way to organize tables by room / building / space.
3. Focus on which tasks occur often, and remove friction from repeated tasks.
Reachability changes when using tablets
Small mobile devices had to be designed for (mostly) one-handed interactions. However, with larger devices, such as iPads, we found that we needed to make use of all available space so that our end users (servers) could benefit from a two-handed interaction, with added white space and clarity, while also providing plenty of space for common tasks to be repeatable and simple.
Patrons think in terms of individual delivery, while the Kitchen thinks of terms of "table" delivery
Something I hadn't anticipated was that the "kitchen" (delivery persona) had to take orders in terms of tables. This taught me the following lifecycle of a table:
1. Table order "submitted" state
2. Table order in a "delivered" and "not delivered" state
3. Table order in a "closed" state
A patron's dissatisfaction climbed significantly if their delivery was delayed in relation to those at their table
What was difficult about this was that a patron's satisfaction or experience was not in relation to the overall table, but the kitchen's delivery was limited to one table at a time. While a single patron could be satisfied when their order was delivered, the experience of the remaining patrons at a "not delivered" table was influenced dramatically based on whether their delivery was delayed in relation to all other patrons.
For Servers, custom orders significantly increased table submission times
Custom orders caused servers to dive deeper into in the interfaces, which caused a level of complexity that increased order times. This complexity followed Tesler's Law, which states "for any system there is a certain amount of complexity that cannot be reduced." Custom orders inherently caused further interactions, slowing the submission of the entire table.
