Not So Hungry Demand Curves
Yesterday, we shared some insight into our long tail of merchants and what the distribution of $40 million in commerce we drove to them last month looks like on the Postmates platform.
Another interesting way to slice this data is by looking at the sales distribution across merchant categories. Prepared food is the highest frequency category, generating the most sales — this makes sense since people eat 3x per day, but other categories are growing significantly.
Retail now makes up 26% of all sales and groceries are 6% of our total sales volume.
Food is to Postmates what books were to Amazon.
We believe there is a paradigm shift taking place in how retail-related commerce is taking place: local commerce is moving online, and on-demand. To that extent, we’ve been actively working on improving our retail category for some time now and over the next few weeks we will announce new partnerships and integrations that will make it even easier for you to buy the things that you want and have them delivered in minutes.
Our desire to offer and grow specific categories inside Postmates is part of our strategy. We love controlling the shape of our demand. Prepared food is ordered mainly during lunch and dinner, but other categories show very different patterns (keep reading). Prepared food is what got our growth flywheel started. Food is to Postmates what books were to Amazon.
We see full utilization of the Postmates platform during peak meal times, but lower utilization during non-peak times.
Peak times present challenges and opportunities. For example, the density of drivers and riders is the highest during peak times, resulting in a more efficient platform. On the other hand, these are also the times that customer service and merchant operations have to deal with extremely high order volumes during a relatively short period of time, as well as delays at merchants and other unforeseeable issues that can have ripple effects throughout the platform. For context, around 63% of our volume happens during peak times.
To complement the demand curve of prepared food, we slowly started to actively promote use cases and categories with different demand curves. The added benefit is the fact that we can start to generate high income for our fleet at any moment in time.
Let’s have a look at a few more demand curves:
This works on the merchant level too. Here is a closer look at the times people order items from Walgreens:
Our data allows us to plot these demand curves for any sort of vertical but their effect on the platform is best shown when combined:
Doing more than food is not an easy task. Some people call it a distraction but if we truly believe in our vision to become the on-demand delivery infrastructure for major metropolitan areas, we have to understand and approach these challenges head on.
If you appreciate difficult, product, engineering, data and marketing challenges come work with us.
My email is: firstname.lastname@example.org
Thank you to Sean Plaice, Kristin Schaefer and April Conyers for their contributions and Jeremy Sypniewski for these hand drawn graphs, based on real data.