Venmo Trends — Uber or Lyft? What is the top Emoji?
The popular payment app, Venmo, allows users to send money from their bank account to another users for free. Venmo payments are public by default (users can opt to hide payments but many, as you will see, are left public) meaning users can see every time another user pays someone. They don’t see the amount, but rather a message indicating what the payment was for. The Venmo feed is overflowing with payments as the service processed over $2.1 billion dollars in one of their recent quarterly reports. I wanted to know what people were paying for so I gathered 1,000,000 transactions (using their public API) between 4/22/2016 and 4/26/2016. I then sifted through the data to find any interesting patterns and trends. All time related data is recorded in GMT time as Venmo does not disclose what time zone the payment was sent from. It is widely assumed many of Venmo users are in the United States, so on the x-axis of many of the charts below I have put the time in the format “Eastern Time/Pacific Time” to give the best idea of what time of day users are paying for certain things.
The first thing I did was look at transactions over time. The chart I created showed an oddly smooth drop-off as you progress from midnight to the morning hours. Based on the near flat slope of the line between 11AM and 5PM ET it would appear users equally likely to pay each other during this timeframe. At first glance this seems intuitive as many users would probably be at work. However, some of this data was collected during a weekend and we can not tell how many users are on “normal” work schedules. It is no shock to me activity heats up at night as people begin to pay each other for drinks, dinners, etc…
I broke out every word/emoji used in a payment message and totaled all of them up for the entire dataset. The first interesting trend I spotted was the difference between the amount of times that Uber was mentioned over Lyft or cab/taxi (the three categories included all mentions of relevant terms related such as ubers, lyfttt, or taxis.). The chart shows how big of a difference (nearly 5x) Uber is mentioned over Lyft or cabs. Although payment messages can sometimes be nothing related to what the payment actually for, it would seem odd that in this case people would be lying about which ridesharing service they were using. From this we can see that Venmo users prefer Uber over Lyft/cab by an extreme amount.
We see that ridesharing mentions have a similar activity shape to total overall payments. The three charts are fairly similar and this shouldn’t be a shock as all provide the same service.
The above chart shows how mentions of ride sharing companies relate to overall mentions at any hour of the day. The chart shows a major uptick in activity during the early hours of the morning in both time zones. This would make sense to me because a ride home seems like the most normal thing to pay someone for during those hours.
This chart compares mentions of a ridesharing company to that individual ridesharing companies overall mentions by time. We can see that people who we believe to be taking a cab charge people at a higher rate in the evening-early morning relative to their total mention count than the Uber or Lyft data shows. This is interesting to me because maybe when you take a cab an individual is more likely to “feel” the effect of paying (resulting in them wanting to get the money from those sharing the taxi with them quicker) than when taking an Uber. I think I would agree with this because when I take an Uber I don’t ever think about how much it costs, I just get out of the vehicle and go about my business. When I take a cab the thought of money is forced into my face and it is apparent that I need to charge the people in the cab that didn’t have cash on them to split the fare. Also, with an Uber/Lyft you have a receipt of how much your ride cost on your phone at all times allowing you to charge people accurately later.
This chart contains certain Venmo terms that I found interesting or odd. I found cash interesting because it shows how payment apps have turned humans into peer-to-peer ATM’s. It would probably be a correct assumption that Drugs are not actually drug related transactions. However, you can assume it would be much easier to sell anything when all you need is an app not a large quantity of cash on hand. Amazon showing up so many times probably indicates the sharing of Amazon Prime memberships in my opinion.
While doing this study I found that many payment messages were just emojis. There were some pretty elaborate combinations but to make this chart I broke out the top 9 emojis by mention. Nothing on this list should shock anyone as some of them show up as suggestions and others are just common things people may pay each other for, like pizza and beer.
This graph covers what sort of food Venmo users are paying each other for. Pizza and beer are probably a no-brainer considering the common perception of Venmo users is that they are of a younger age demographic. However, it looks like a great time to start up a Thai or brunch spot. By collecting more data over time trends could be uncovered about what people are eating at different times of the week or year.
The date that I pulled this data was during Coachella time, so the fact that this shows up is not surprising. This also gives you insight into who Venmo users are if you have any idea what Coachella is. Vegas being popular might also give you more knowledge about the average Venmo user.
The most important thing that I got from looking at all this data is that there are trends people follow and the data is available in unusual places. By looking at stuff like this we can understand how people exchange money and what they exchange it for. When it comes to ridesharing you hear in the news of a battle between Uber and Lyft. When consulting this “third party” data source it is obvious by simply looking at a chart: Uber is dominating. I was also somewhat shocked Thai food was mentioned so often but as a Venmo user myself, I do eat a decent amount of Thai food. None of this data can be guaranteed to be accurate but it gives us a look into what people are paying for and when they are paying for it.
Hayden
*Charts created using Highcharts
