Back in October, we added dockless bikeshare to our app, Transit. With new bikeshare operators popping up all across America, we wanted to make it easier to find the nearest bike without switching between multiple apps.
We started asking ourselves some questions: Are Transit users embracing dockless bikeshare? How many bikes are actually available? Is it easier to find a docked or dockless bike? Is dockless serving rich neighborhoods and poor neighborhoods equitably? And what did we learn about how bikeshare can become even easier to use?
Transit users ♥ dockless bikes
The first thing we did was look at the share of our users who use dockless bikeshare.
In the Seattle area, 13.5 percent of Transit users have a dockless app installed. In the DC area, where dockless competes with Capital Bikeshare, 4 percent of Transit users have downloaded a dockless app. (While there’s room for growth, dockless is already ahead of car2go, which is used by 3 percent our DC-area users.)
In both cities, half of our users with dockless ended up downloading more than one company’s app, suggesting that one provider isn’t enough. In fact, using Transit to look for a bike makes you much more likely to find a ride than if you rely on any single dockless app.
Here’s how we know: We took an anonymous sample of more than 5,000 app launches in DC from October to January, and pulled the locations of bikes within a quarter-mile. Since Transit is used more frequently in dense areas and during peak commuting hours, the places where our users are opening the app is a good proxy for where bikes are in demand.
We found that relying on an average individual dockless app leaves you with a less than 50 percent chance of finding a bike within a quarter-mile, or a five-minute walk. Aggregating all the dockless bikes together brings that number above 80 percent. By combining Capital Bikeshare and dockless in Transit, your likelihood of finding a bike soars to 91 percent.
Integrating bikeshare alongside other transportation choices doesn’t hurt either. In October, the first full month of dockless in DC, our users tapped dockless options thousands of times. Our users even chose dockless trip plans more than twice as often as Capital Bikeshare.
But we didn’t just compare which bikes our users tapped. We looked at the entire flow of bikes during a three-month period after five dockless bikeshare companies launched in DC from October 13, 2017 to January 13, 2018. (Not exactly the warmest time of year for biking in DC, but hey, at least we got October in there.)
We were able to reach some surprising conclusions about the difference between docked and dockless bikes.
How many bikes are there?
What we did next was pretty straightforward: we started to count bikes.
Before we keep going, there are a few things to keep in mind. We only tracked the number of bikes shown as available in each bikeshare provider’s API. If a bike was turned “off” (because it was already rolling across town or getting repaired in the shop), we didn’t count it. Plus, we only counted bikes we could spot within the District of Columbia. If a bike was out in Virginia or Maryland, we excluded it from our numbers.
To start, we looked at the number of bikes at 5 AM each day. Since very few people are awake and riding bikes at that hour, we bet that almost all the bikes on the street would show up in our numbers.
On most days, there are about 1,100 dockless bikes on the streets of DC.
That’s a lot of bikes! But Capital Bikeshare, the dock-based system that launched in 2010, is still the big dog of DC bikeshare. In October, it logged nearly six times more trips than all the dockless operators combined, according to the District government — 338,152 trips by CaBi, compared to 56,477 on dockless bikes.
We saw a lot more Capital Bikeshare bikes, too. Most days, when compared to their dockless cousins, there were about twice as many Capital Bikeshare bikes available. (It's worth noting that under DC's current regulations, each dockless company is limited to 400 bikes at any one time.)
When you combine these two stats — the number of trips and the number of bikes — you get one of the best bikeshare measuring sticks: how many trips a bike makes in a typical day. The higher the number, the more popular the system. To calculate this, we looked at two stats for an average October day: the number of trips (1,822 dockless, versus 10,908 CaBi, according to DC government) and the number of bikes available (1,146 dockless, versus 2,056 CaBi).
Our conclusion? We estimate a typical dockless bike was ridden 1.6 times each day in October, whereas a Capital Bikeshare two-wheeler clocked 5.3 daily trips. That made Capital Bikeshare three times more popular that month than its dockless cousins. But don’t feel bad, dockless fans: that’s pretty solid for what was, at the time, a brand-new service. (And in Seattle, dockless ridership has eclipsed the city’s previous dock-based system.)
Where are the bikes?
So, where in DC are these bikes, exactly?
There’s been a lot of discussion about making sure bikeshare is serving all communities — not just the wealthy few. So we set out to answer one question as part of that larger puzzle. We wanted to know: are dockless bikes equitably distributed east of the Anacostia River compared to Capital Bikeshare?
Here’s how we figured this out: Every 10 minutes, we checked how many bikes were available in each ward. If, for some reason, there was no data from even one bikeshare operator at a given moment, we chucked all of the stats from that point in time. (It ended up not being a big deal: we only had to chuck out five percent of the data.)
Even with the numbers in hand, counting bikes is tricky because the flow of bikes changes constantly. On weekdays, people tend to ride downtown in the morning and back home at night. We couldn’t just measure each ward at a single point in time every day if we really wanted to understand how bikes are distributed across the city. So we looked at all of the points in time — by calculating the median number of bikes available in each ward, using all the checks we made every 10 minutes.
How does dockless fare when compared with CaBi? Turns out, it’s a draw.
We found that 11 percent of bikes, whether docked or dockless, are in Wards 7 and 8. For both types of systems, more than 70 percent of bikes are in Wards 1, 2 and 6, covering downtown and dense nearby neighborhoods. And fewer than 20 percent of bikes are in Wards 3, 4, and 5, which cover mostly outlying neighborhoods west of the river.
There are a lot of initiatives focused on improving equity in bikeshare, and our number-crunching looks at just one part of the picture. But when it comes to comparing the distribution of bikes in DC’s lowest-income wards, Capital Bikeshare and dockless bikes come out about the same.
When can you find a bike?
If you’re in need of a bike, when do you have the best chance of finding one? We looked at two popular locations as examples.
First, we looked at the average number of bikes available near downtown’s busy Gallery Pl-Chinatown Metro station, by finding all the bikes within a quarter-mile radius of 7th Street and G Street NW.
There’s five Capital Bikeshare stations nearby, with a total 128 docks. On weekdays, the average number of bikes in those docks rises from its overnight low, accelerating during the morning rush to peak around lunchtime. There’s another small spike around 7 PM as people come to Chinatown for nightlife. In contrast, the number of dockless bikes holds steady.
A couple miles to the north is Columbia Heights, a popular residential and commercial neighborhood. There, five Capital Bikeshare stations sit within a quarter-mile of the Metro, with a total 109 docks. The average number of Capital Bikeshare bikes available plummets during the morning rush, then starts to rise again as people return home. But the number of dockless bikes remains remains relatively steady… just like downtown.
It’s a pattern we saw again and again, unsurprising to any regular DC bikeshare user: busy locations saw swings in the number of Capital Bikeshare bikes, especially during rush hour, but less variation in the number of dockless bikes. On weekends, the number of bikes (both CaBi and dockless) remained a bit steadier at locations throughout DC.
What explains these different patterns? It could be because (as we showed earlier in the post) there are fewer trips by dockless bikes relative to the number of bikes on the street, leaving most of them untouched. After all, dockless bikeshare is a pretty new service — they’re still in the early days of growth. It could also be because Capital Bikeshare (with its annual payment option) is more popular for daily commuters, whereas dockless relies on a stream of pay-per-ride users throughout the day. It could be a combination of these and other factors.
Having all the data in one place is good for everyone
There’s a lot of exciting experimentation going on with dockless bikes, e-bikes, and now scooters. We hope these services will become even more popular, reducing car usage for short trips and getting more people to connect directly to transit… sometimes, too directly.
Putting these new services into one app is great for everyone: riders, operators, and cities.
For riders, it’s better to launch one app with a 90 percent chance of finding a ride — instead of opening six separate apps that are each less likely to turn up a nearby bike. But aggregating location information isn’t enough. Riders want to easily sign up, pay for and unlock bikes without having to download all those different apps in the first place.
As for operators? Transit can help drive trips to these operators that otherwise wouldn’t have happened if users had to discover each service on their own.
Consider this: in October, when there were just 56,000 dockless trips in DC, Transit users tapped to download or launch a dockless app over 3,700 times. Our users also chose dockless options in our trip planner 2,700 times — more than twice often as Capital Bikeshare trips. (This, despite the fact that Capital Bikeshare logged five times as many trips. Perhaps Transit’s cohort of early adopters are onto something.)
For cities, data from aggregators can provide a richer picture on travel patterns. How are users stitching together their trips? Are they using bikeshare to complement public transit trips, or are they more likely to take bikeshare vs. ridehail under certain conditions? By understanding riders’ multimodal travel patterns, cities can create better-informed transportation policy, and ultimately, deliver better service to their citizens. Alongside contributions from indvidual operators, like Uber’s Movement platform or LimeBike’s Year End Report, aggregation data can give cities a comprehensive view on how different modes and services are used. Our DC bikeshare research is just the start.
As we dig further into the numbers — and forge closer partnerships with operators and cities — we’re thrilled to be getting closer to a “big picture” understanding of how people get from a to b.
A version of this article was originally published on Medium.