Taking Stock of Dockless Bike-Share in Seattle

Photo: City of Seattle
Photo: City of Seattle

The most interesting experiment with dockless bike-share in an American city right now is happening in Seattle. After its publicly-funded bike-share system, Pronto, went kaput without ever scaling up to significant size, last year three bike-share start-ups — Lime, Spin, and Ofo — introduced a combined fleet of 10,000 bikes in the city.

Seattle and Dallas are the only two American cities with dockless bike-share fleets of at least 10,000 bicycles. And only Seattle has the beginnings of a decent bike network and relatively high bicycling rates for the U.S. For these reasons, it’s an interesting test case for the dockless bike-share services.

With the City Council preparing to grant long-term permits to the companies, Seattle DOT put together a summary of how they’ve performed so far [PDF].

Unfortunately, the trip data only covers the period from July through December. That raises some red flags. Bike-share trip data is easy to compile and share, so why are the five most recent months not included?

The available data shows that by their sheer volume of bikes, the dockless services are outperforming the tiny Pronto system. And in a survey, about a third of Seattle residents said they had given the bikes a try, which held steady regardless of race.

But the bikes aren’t used as intensely as station-based bike-share systems of equivalent size. The average bike-share bike in Seattle is used for 0.85 trips each day. That’s less than half the rate of Chicago’s Divvy, which gets more trips out of 5,800 bikes than Seattle is getting out of 10,000 bikes. New York’s Citi Bike, with 12,000 bikes, produced an average of 51,800 daily trips from July through December, more than five times the rate in Seattle.

Still, compared to Pronto and other small station-based systems — which is all most American cities have been able to muster — the dockless services are doing well. Patrick Taylor at the Urbanist writes that he’s come to rely on them:

As an frequent bike rider who lives in Southeast Seattle, works in Central Seattle and spend a fair amount of time in North Seattle I have seen the growth of floating bikeshare in all parts of the city and witnessed it becoming an integral part of my own transportation. One of the most satisfying things for me to see is how often it is used by people in my neighborhood of Othello, which is not the most bike-friendly part of the city. I just opened the Lime app and there are seven bikes within a few blocks of me, including three e-bikes. In my experience it seems like it has increased the number of people on bikes and the types of people (by which I mean not spandex warriors, Day-Glo commuters, bike punks, or the usual suspects).

In China, dockless bike-share companies have increased bicycling mode share by flooding cities with hundreds of thousands of bikes. Seattle isn’t close to that volume, and the city’s bike-share manager told the Seattle Times in December that the companies aren’t looking to expand much more due to the cost of maintaining larger fleets.

But if the dockless fleets do get bigger, survey data suggests the public will welcome it. Despite all the headlines about bike clutter, 74 percent of survey respondents said they have a favorable opinion of the services.

Another interesting takeaway: Few people are wearing helmets when they use bike-share, despite Seattle’s helmet law. The city reports that only 24 percent of bike-share riders used helmets, and preliminary data from a University of Washington study shows no increase in head injuries.

  • Larry Littlefield

    I suspect that Seattle doesn’t have as many barriers to bike ownership — parking, theft — as NYC. Thus those trying bike share are more likely to buy their own if they enjoy traveling by bike.

  • Scott

    Just a quick note, the helmet law is actually a King County provision. https://www.injurytriallawyer.com/faqs/is-there-a-helmet-law-in-seattle.cfm

  • Alec

    Just got back from Seattle. The Lime Bikes are janky AF – love the idea of the dockless bike but every other bike was broken in some way or wouldn’t unlock. Made me really appreciate Citibike.

  • meelar2

    The lack of trip intensity per bike doesn’t really bother me–after all, the 10,000th bike in a smaller city like Seattle is way more marginal than the 5800th bike in Chicago.

  • Fish

    We have different barriers (theft is definitely a huge one we share)..they are called hills and only 1 legit train line. Biking up the hill is a barrier for a lot of people but taking public transit up a hill and then a bikeshare down works well. Additionally, many people are using the bikes to get to our 1 train line or several “rapid” bus lines that are beyond walking distance for most people but work well with a bikeshare.

  • david reeves

    Thanks for this article. I’m all for rigorously taking stock of bikeshare. But a few comments:

    1) Why focus on rides per bike as the main metric for success? From the city’s perspective, it’s largely irrelevant; what the city should care about is total number of trips. And in a dockess system with low-cost bikes, it’s a good thing to have more bikes than you need, because the availability of open docking spaces doesn’t matter anymore. Let’s look at the density of bikes relative to the population:
    Seattle has one bike for every 70 residents;
    NYC has one bike for every 700 new yorkers.
    Chicago has one bike for every 482 chicagoans.

    Is it at all a surprise that rides per bike is higher in cities with four and 10 times the population, with a similar number of bikes?

    2) how did the 1st 6mo of bike share in Seattle compare to the first six months in Chicago in 2013? Or the Have to compare apples to apples, and habit change takes time. Comparing a service that just launched, and that didn’t even have a full fleet initially to a mature service 5 years in is not an apples to apples comparison.

    So viewed per capita, seattle’s system is less used than NYC citibike in the six months after launch in 2013, (.66 rides per resident vs .73 trips in NYC), but far ahead of chicago, which was at .27 rides per capita. That’s not even considering that the seattle system launched with fewer bikes during the peak cycling months in the summer.

    3) In what way could the July-December study period possibly be a red flag? It was a pre-planned evaluation period for the pilot project, consolidated from three private operators, with an analysis done by a private body (the UW), and combined with other survey data. All of this s described very clearly on Seattle DOT’s website. You’re reading the report of the evaluation period. The city is generally very open about providing real-time data.

    Other things that would be worth mentioning:
    1) hills. Not sure if you’ve been to seattle, but some of the most popular destinations are 400ft vertical climbs in under half a mile. Brutal on bikeshare bikes.
    2) The recent availability of e-bikes (after the study period)

    lastly: I’m curious why streetsblog seems so consistently negative and skeptical about dockless systems. Why is that?

  • mx

    We’ve recently added a bunch of bike share e-bikes (both docked and dockless) in San Francisco, and they’re a revelation when they come to the hills, especially since pedaling 40lb non-electric bike share bikes up hills is a pretty awful experience. They’re not going to make getting up a 15% grade entirely effortless, but with some diligent route selection and/or pedal power, it opens up a lot more areas to those of us who aren’t in superb shape. I’d think they would be a good fit for Seattle as well.

  • JacobEPeters

    First off to address the last question. Read any of the streetsblog chicago articles about dockless bike share to see that it’s not consistent negativity. I don’t even think that this is a negative article about dockless bike share. Skepticism is warranted when private companies could leave at a moments notice & eliminate the entire bike share network. Skepticism isn’t opposition, & criticisms are helpful feedback for any maturing industry.

    1) The reason for focusing on rides per bike is because bike share derives its efficiency from providing more rides per day than a personal bike. It becomes less relevant to cities with dockless systems because they don’t pay upfront for the bikes, but it becomes very relevant to the dockless bikeshare providers because they want to maximize revenue & low utilization means that their overhead is too high.

    In both New York & Chicago the docked systems do not cover the entire city. Which is why they have fewer bikes per resident but a more efficient utilization, it isn’t as much about how many bikes per resident it’s about how many bikes per square mile. CitiBike has only ~ 30 square miles of service area, which means that it has a much higher density of bikes than Seattle. That being said the Divvy coverage area is larger than the entire city of Seattle, so hopefully as the Seattle program matures it will see increasing utilization from a higher density of bikes than exist in the Divvy service area.

    2) The first 6 months of Chicago’s Divvy program involved scaling up from 750 bikes to 3,000 bikes. There will never be an apples to apples comparison, because every launch is different (which is why different cities will need different combinations of docked & dockless services). That being said, it seems that docked bike share programs benefit from 2 characteristics. A) they tend to be located in the densest parts of cities, where dedicating street space to stations is justified by a high density of destinations & therefore high usage rates B) the reliability of knowing where a station is, reduces the barrier of uncertainty that can discourage a potential user from planning their trip around using a shared bike. These 2 aspects probably explain why a docked system requires fewer bikes to serve as many trips as a dockless system.

    3) The author seems to be saying that the 6 month study period might not be enough information to find conclusions when there are almost 6 months of additional data that have been generated since then.

    I’ll trade you the pedal assist Lime bikes that have recently been dispatched on the south side of Chicago (& cost $.15/min) for the regular Lime bikes that have a flat rental fee.

  • david reeves

    Jacob, really appreciate the thoughtful response.

    Completely agree that healthy skepticism and constructive criticism are always appropriate.

    Regarding your first point #1; I don’t believe personal bikes are the relevant comparison with bikeshare. It’s a new mode, and done right, should only have minor overlap with personal bikes. Its competition is walking, rideshare, public transit, and personal bikes, in roughly that order. And arguably it makes public transit trips more feasible because the last mile to your destination is much faster. Rides per bike matters a lot to the profitability of the bikeshare companies, but, I submit, not at all to the city or to its inhabitants, all of whom benefit from a greater total number of rides.

    If you optimized for rides per bike, you’d err on the side of providing bikeshare to fewer people, at the cost of fewer overall rides in the city.

    On the other hand, focusing on overall rides results in investing more in supply ahead of demand to create the *experience* of completely ubiquitous bikes, everywhere. This is the lesson of Uber – that they heavily subsidized supply in their early stages so that a car would always be around the corner, even before rider demand caught up.

    #2 – I’m not interested in a competition between cities, honestly. I just know (see above) that building demand takes time through habit change and repeated exposure. So a comparison between a service that’s had 5 years to market itself with one that just launched isn’t really fair, unless you caveat your analysis that way.

    #3, regarding the period of data – your statement, which I agree with—is much milder than the term “red flag” which suggests to me either high danger or some kind of willful misrepresentation. Angie made no mention of the pilot project, the pre-planned study period, etc.

    So – I accept what you say about skepticism. I guess what I was hoping for was a process of taking stock that felt a little more accurate and reflective of the nature of launching a new service and the wildly different population densities. Thanks for listening.

  • david reeves

    Hey, streetsbloggers – a comment I made was flagged as spam, I assume by accident — would you mind restoring it? Thanks for a great site.

  • cjstephens

    It’s good to be reminded that some pilot projects are just that, a test run to see what works and what doesn’t work. Smart cities can learn as much from a bike share test that fails as from one that is a success. Negative headlines can be discouraging, but they also can let the bike share systems know what other people see as priorities. Let’s keep testing to see what is the best solution for every city. One size probably won’t fit all.

  • david reeves

    Jacob, really appreciate the thoughtful response.

    Completely agree that healthy skepticism and constructive criticism are always appropriate.

    Regarding your first point #1; I don’t believe personal bikes are the relevant comparison with bikeshare. It’s a new mode, and done right, should only have minor overlap with personal bikes. Its competition is walking, rideshare, public transit, and personal bikes, in roughly that order.

    Bikeshare “derives its efficiency” from stealing share from other less efficient modes, and from enabling trips that wouldn’t otherwise have happened at all.

    Focusing on overall rides results in investing more in supply ahead of demand to create the *experience* of completely ubiquitous bikes, everywhere. This is the lesson of Uber, which heavily subsidized supply in their early stages so that a car would always be around the corner, even before rider demand caught up. This is a more promising approach to grow faster; but it does intentionally hurt metrics like rides per bike in the short term.

    If you optimized for rides per bike, you’d err on the side of providing bikeshare to fewer people, at the cost of fewer overall rides in the city.

    #2 – I’m not interested in a competition between cities, honestly. I just know (see above) that building demand takes time through habit change and repeated exposure. So a comparison between a service that’s had 5 years to market itself with one that just launched isn’t really fair, unless you caveat your analysis that way. And population density matters, not just density of bikes.

    #3, regarding the period of data, your statement—which I agree with—is much milder than the term “red flag” which suggests to me either high danger or some kind of willful misrepresentation. Angie made no mention of the pilot project, the pre-planned study period, etc.

    So – I accept what you say about skepticism. I guess what I was hoping for was a process of taking stock that felt a little more accurate, constructively critical and reflective of the nature of launching a new service and the wildly different population densities. Thanks for listening.

  • david reeves

    Spot on. They’re a game changer here. Excited to see scooters too, once our enlightened city government sees fit to allow them.

ALSO ON STREETSBLOG