Since it's one of the hottest topic out there at the time of writing (and I think we could be of great help being people who live in close contact with data), I chose three KPIs somehow related to ESG concepts.
- Top performing sustainable store/restaurant types by volume of orders across different geographical areas
- #Accidents and injuries suffered by couriers
- Various pre-order metrics, with the ultimate goal of measuring the effect that better supply processes by stores have on food waste
This I guess is quite difficult to quantify, as a result of two different conditions:
- Not all accidents/injuries are reported to Glovo, especially the less severe ones (non-serious falls from the bike, light damage to vehicles, etc.)
- Not all injuries come as direct consequence of a specific circumstance/event (let's imagine a courier who's suffering a serious throat inflammation due to the repeated travel through a chemical district to make his deliveries)
First action point then might fall around the ever-present question "Where's the data?". In other words, how do we incentivize couriers to provide us with fast, accurate, timed and geolocated evidences even for those minor ones we're missing today?
A good starting point might be the technical answer (of course the incentive has to be built on a different driver), a simple "take a picture and report to Glovo" feature in Courier app. Being the photo timestamped and geolocated we can easily match its metadata with the courier's in progress delivery, also double checking his historical records to prevent most obvious frauds and duplicate claims. Talking about incentives, we can think of a system that awards accurate reports over time with a positive effect on that courier's overall rating (why not a new bike someday). Open data as well can be of great help in gaining crytical data mass and ultimately even a better statistical significance in our dataset (high crime city areas, most frequent road accidents areas, etc.).
But what is the ultimate purpose of having this system in place? And how do we manage to operate this new data to create safer working conditions for our couriers?
One possible use could consist in the systematic correlation of these new data points with the component responsible to provide the delivery suggested optimal route to couriers (in case you have one already in place): by suggesting safer routes there's a clear chance we can obtain a happier workforce, save money on assurance costs, earn a more sustainable profile as an employer and even improving some other time related KPIs (safer -> faster).