Wait Times on Google
Get estimated wait times for over 1M dining establishments
Average wait is very useful to users planning for travel, medical & especially dining
Today, we can confidently overlay usual wait times with live and historical crowd volume. Right now, we can surface this info for over 1B dining establishments. We considred a number of use cases and existing models to help determine the right UI and location to start with.
Existing data viz patterns
Existing Placesheet modules
Research + use cases
Since we have the data, we focused on immediate & long term dining use cases
We knew that we had the data to infer a mean wait time average for over 85% of restaurants listed on Google. Other verticals needed higher confidence rates to be considered.
What's the longest wait next Tuesday?
Is there usually a wait at my local cafe?
Explorations + findings
Pairing with date & times was more intuitive than with hours of operation
The best place to surfce this was our existing Popular Times Graph. A slider was too complex for v1, thus selecting an hour manually was our first approach. Though the tooltip became quite verbose, depending on time and language, we felt it was a good place to start. Our existing live snippet didn't localize well and wasn't a noticeable location when traversing hours. Enter: the tooltip.
UI + placement explorations
Evolving snippets into tooltips + summaries
Through prototyping and a few user studies, we found that elements with shadows fit too closely with our action/tappable langugages and needed to be toned back. Alongside copy writing and typeface style, we also experimented with tooltip carets, however it proved problematic rendering for places that stayed open 24 hours.
An interactive tooltip allowed for a more flexible UI
2/5 restaurants have varied wait times between mornings, afternoons and evenings. A tooltip UI allowed users to traverse hours and showcase a peak wait time for the selected day. Originally, the beak on the tooltip aimed to add more attention to the selected hour, but didn't scale to the edge well. The summary focused on the worst case scenario for that day. Most users wanted to get a general sense for when not to go, but also wanted the ability to see other times.
Interaction + visual design
Heirarchy needs improvement as we expand to new verticals
Even with an exceptional and accelerating growth rate, the v1 moved pretty quickly and didn't account for everthing we wanted. However, impact was high and we are already working towards feature improvements and better localization. This includes partner and vertical expansion, frictionless payments and personalized views of your transactions.