6 ms·
Data on Uber Usage in the U.S.
- minimaxir 11y agoThis is not a very insightful analysis, which is surprising since I assume it's coming from a YC batch company. (given that katm submitted it). The stacked area charts specially are very confusing to read and are impossible to compare the changes over time of the different categories. There also is no sample size given, which is relevant since the average values are close and therefore we can't easily say if the average values by subgroups are statistically significant at a glance. (I assume it's high, but it needs to be explicitly stated) The title is also unanswered in the article. "More than You think" compared to what? Normal car/taxi traffic? All the data shows is that people ride Uber a lot.
- mikebabineau 11y agoThanks for the feedback -- we'll definitely be improving these over time! To answer the question on sample size: This analysis was done across >300,000 unique Uber customers from a very large, representative sample of US consumers.
- mikebabineau 11y agoMike and Lillian from Second Measure here. We're just ramping up on these sorts of posts, so we'd love to get feedback and answer any questions!
- JoshTriplett 11y agoOne thing that struck me as odd from the data: bookings are growing just below customers, on the same scale, which seems inconsistent with only ~30+% of customers riding only once per month. The only obvious way I can think of for that to happen is if there are many customers who ride less than once a month, which would actually make sense; I use Uber a few times a year, when I travel, but I don't use it locally. So, it seems like the "Domestic Customers by Monthly Rides" graph needs to have a "0" included in it (and then a second graph without it if that squashes the rest of the graph).
- lillian 11y agoThanks for the comment! It'd definitely be interesting to see this same analysis for annual rides to see how many people fall into a similar category as you. To quickly clarify: The growth chart only includes customers who paid for an Uber ride in that month, so any customers riding 0 times are not counted. The "Domestic Customers by Monthly Rides" graph simply takes that set of customers and breaks them down by # of rides in that month. Bookings can grow more slowly than customers if customers are spending less on average. This is in fact the case - in July 2012, customers were spending ~$85 a month on Uber. This has dropped steadily over the last three years to ~$70 a month. This makes sense given that Uber has lowered its average price over the years, with the introduction of UberX and UberPool.
- fraserharris 11y agoCan you get geographic data with these purchases too? What % of people who don't have local Uber service use Uber when they are travelling? It would point towards market expansion markets.
- lillian 11y agoThat's a great point and something we'll be adding. It'd also be a good way to see how different people use Uber (commuters vs. travelers) - even if Uber is available locally.
- deleted 11y ago
- FamousAspect 11y ago"Interestingly, highly engaged riders (those averaging 12 or more rides per month) spend less per ride than infrequent riders" Thinking about my own Lyft use here, I became a frequent user after discovering that Lyft Line was a great and cheap way to get to work if I was running late. So one could hypothesize that people become heavy users because Uber allows them to take a frequent trip affordably.
- lillian 11y agoI agree - the more frequent riders are likely using Uber for shorter distances, like commuting to work. It's great to hear the other side of the numbers (riders' anecdotal experiences) so thanks for sharing!
- dang 11y agoWe changed the title from "People Ride Uber More Than You Think" to get rid of the linkbait "you". Happy to change it again if anyone suggests something better.
- lillian 11y agoThanks dang. The linkbait effect was unintentional - we'll be more mindful of this going forward!
- compumike 11y agoThis looks like an amazing dataset. Are you able to extrapolate from your sample up to their full US revenue? How often is your dataset updated / how much lag is there? Looking forward to your future blog posts!
- lillian 11y agoYes - when a company reports their actual numbers, we can get a sense of what % of revenue we see and extrapolate to full US revenue. We update our data weekly - our blog too so stay tuned!
- qq66 11y agoWhere does Second Measure capture this data from? Are my Uber rides in this study?
- lillian 11y agoWe analyze spending across a large number of US consumers. It’s anonymized, so we couldn’t tell either way!
- thedogeye 11y agoHow can a company opt out of your dataset?