The bulk part of shared mobility apps (like those of electric scooter and car-sharing companies) display similar home screens when launching their apps; a map pops up, and in most cases, this also triggers a search of nearby vehicles. “App-open” events are often embraced as trustworthy indicators of actual demand amongst mobility service providers, but Zoba argues that these data are typically subject to lots of noise and confounding factors.
They give three explanations of why these perceptions are often flawed:
1. People search without wanting to reserve vehicles. Some app-opens trigger searches even though there is no actual demand, the user might only want to investigate current pricing, availability, review past rides, update profile, or just happen to press the app icon accidentally to mention a few. Moreover, some in-app actions often trigger new searches, like when a user drag and zoom the map to find vehicles in other areas.
2. Demand might be present without any search being made. For instance, some people hope to find an available ride on their way to work. If the supply is insufficient in their neighbourhood, they might give up searching and the mobility service provider will end up failing to capture demand where it actually exists.
3. Searches that are actually triggered by users who want to reserve a vehicle could be difficult to interpret for several reasons. A user might spot a vehicle on the street and spontaneously decide to use it, search for vehicles in other locations due to lack of available rides at their present location, or search for vehicles at more convenient landmarks. All of which can form a misleading picture of spatial demand.
An important finishing point from Zola is that searching is cheap. It costs nothing and requires minimal time and effort. Therefore, users often trigger a lot of searches before deciding on what vehicle to reserve, where and when.
The “right way” of predicting demand is a tricky question. I think all three explanations highlights important factors for mobility service providers to consider. Yes, users might generate searches unintentionally due to the app-functionality, but at the same time, it’s a necessary service to maintain for people who actually want to reserve a vehicle. Regarding the demand without any search being made or the uncertainty of spatial demand, an idea might be to involve people in the process to a greater extent by allowing them to request new supply at certain areas or to somehow mediate where they actually wanted to travel from.
Written by Hampus Alfredsson, RISE Viktoria.
1. 2019-08-14. App-opens: a perilous proxy for shared mobility demand.