Washington Post – "The Business of Bike Sharing"

[mstone]

Thinking outside the box is great, but thinking critically is still important. Somewhere between 500,000 and a million people move around every morning in the cabi service area. If guaranteeing any non-trivial fraction of those people is among the criteria for cabi’s success, it’s doomed to failure. None of the novel rebalancing ideas do more than change the magnitude of the failure by a tiny amount.

That said, I think the idea that cabi should be a commuting tool is mostly wrong. If you want to move a large number of people in the same direction at the same time, you build a heavy rail system. If you need slightly less capacity you build a light rail or a bus. You don’t build a bunch of single passenger vehicles, have the public move them one way, then put them on a truck to take back to their origin–you’d be better off just putting the public on a bus in the first place, with less overhead. Bikeshare’s strength is short errands, last mile completion, etc., throughout the day, especially for low-volume random routes (such that running mass transit on the route would be impractical). Sure it can be useful at peak times also, especially for an oddball case like “my bike is in the shop”. But in no way should anyone try to bill it as a general purpose commuting system for everybody: it just can’t do that thing, and it’s setting the system up for failure.

[PotomacCyclist]

But the fact remains that many local residents will be drawn to join CaBi for commuting purposes. You can’t ignore that, whether or not the system is advertised as such. CaBi clearly recognizes this because of their efforts to minimize the dockblocking problem downtown. If they don’t continue to work at it, enthusiasm for the system can stagnant among local residents. I’m not saying that this is the only vital issue for CaBi as it continues to grow in popularity and coverage. But it’s something they should continue to keep in mind and address.

As far as the 500,000 number, where are you getting that? 500,000 people are not using or attempting to use CaBi to commute downtown. Most people take non-active modes of transportation such as Metro, driving or buses. Some walk. Some ride their own bikes. CaBi certainly does not need to handle 500,000 bike commuters each morning. Not even close.

And who said anything about guaranteeing any specific service? They can simply work on improving the morning service. Just because you can’t do something perfectly doesn’t mean it shouldn’t be done at all. They have already tried the methods I suggested, but in a half-hearted manner (except for the station additions). One method does not need to solve the problem by itself. If it costs relatively little money, then why not do it? It would not be tied to any sort of service guarantee. I don’t understand that thinking.

[mstone]

@PotomacCyclist 53133 wrote:

But the fact remains that many local residents will be drawn to join CaBi for commuting purposes. You can’t ignore that, whether or not the system is advertised as such.[/quote]

You can’t ignore it, but you certainly shouldn’t support it, and I think that they should (continue) to clearly message that it’s not the intended use. The problem will tend to be self-correcting as people learn that it’s not a reliable mechanism and come up with better alternatives. The last thing they should try to do is “fix” the problem because doing so will suck up resources, make the program non-viable, and eventually kill it (ruining it for people who are actually using it as intended).

Quote:

As far as the 500,000 number, where are you getting that? 500,000 people are not using or attempting to use CaBi to commute downtown. Most people take non-active modes of transportation such as Metro, driving or buses. Some walk. Some ride their own bikes. CaBi certainly does not need to handle 500,000 bike commuters each morning. Not even close.

I said “a non-trivial fraction”, not the entire half million. But you can’t ignore the fact that’s the pool of potential users. You’re trying to weasel and say “well, not everybody, just the ones that want to” which is a BS answer. If any non-trivial fraction of the potential userbase attempts this, it won’t work. How do they make sure that doesn’t happen? They limit availability. But you argue that cabi shouldn’t limit availability, but they don’t need to worry about too many people using it, but they don’t let enough people use it now. There’s a logical disconnect buried in there somewhere.

Quote:

And who said anything about guaranteeing any specific service? They can simply work on improving the morning service. Just because you can’t do something perfectly doesn’t mean it shouldn’t be done at all. They have already tried the methods I suggested, but in a half-hearted manner (except for the station additions). One method does not need to solve the problem by itself. If it costs relatively little money, then why not do it? It would not be tied to any sort of service guarantee. I don’t understand that thinking.

Providing non-guaranteed best-effort service is exactly what they’re doing now, and you’re complaining about it. They can implement all the marginal improvements in the world, and they will be open to exactly the same criticisms as now, except that they would have distracted themselves trying to “fix” a problem they can’t solve instead of focusing on their core competencies. Something can cost little money (capital) and still cost an enormous amount of time/energy and generate huge opportunity costs if that’s what they’re wasting their staff time on.

[PotomacCyclist]

The highest weekday total ever has been 10,166 rides, which includes the morning and the evening rush hour as well as all of the other trips throughout the day, in the entire system. Don’t use inflammatory terms like “BS” when your half million number is not a serious consideration. Many people’s commutes are far too long to be taken on CaBi. Many others simply won’t bike no matter what. Others may not live near or work near a bike station. The potential pool of morning users is far, far smaller than 500,000. And even with the 10,000 number, say a third or so are in the morning rush, but that includes trips to all the other places in the CaBi network. While many ride to downtown in the morning, there are also a lot of trips to other neighborhoods and areas, including work centers like Rosslyn-Ballston, Capitol Hill, Dupont Circle, etc. This is important when tossing out numbers because this further decreases the traffic volume under consideration.

They do message that commuting is not the intended use. That’s fine and I think that’s what they should do. But they can also take low-profile measures to ease dockblocking, which they already do. I’m merely suggesting that they can expand on those measures. They don’t need to publicize them, although they don’t need to hide them either.

As far as weaseling out of a number, no, I am not. You honestly think that 500,000 people will bike to work downtown? I can absolutely guarantee that won’t happen in the foreseeable future. You know this and everyone else knows this. It is not BS or weaseling. It’s a simple understanding of American behavior. WABA and other groups promote cycling in all sorts of situations, with the ongoing understanding that you aren’t going to get 100% to bike commute. Or 50%. Or even 30% or 20% in the near future. Even 10% would be a big number and that would include bike commuting to all of the work centers in the area, not just downtown D.C. Yes, these distinctions ARE important. The potential pool of bike commuters to downtown drops far below 500,000. How far below? I don’t know exactly. 1,000? 2,000? And that’s not just to one specific station. There are multiple stations downtown, which drops the per station demands even further. This makes is much more reasonable to engage in low-profile efforts to further ease morning CaBi use.

Will it ever guarantee service at these times and at these locations? Probably not. But it can sure help, and it would not take anything close to the herculean efforts you are implying.

[lordofthemark]

Whatever share of regualar bike commuters to downtown is going to be versus non-active modes and walking, I think Mstone is correct that the best way for most of them to get downtown is going to be to use their own bike. Its got to be cheaper to provide more bike parking than to provide more Cabi docks, and as he says, busing bike back for this purpose doesnt really make sense.

So what can be done

A. Don’t some cities provide incentives for reverse biking – if you go from a dock blocked station to one that needs bikes, you get extra minutes or other incentives? That seems like a desirable way to deal with some balancing issues

(edit – you mentioned that above, I see. I think thats a good idea, and would not be subject to the kinds of problems Mstone fears)

B. Doesn’t the software make it possible to ID who is riding where? Could it be possible (without violating privacy) to somehow use that info to market regular commuters to dockblocked stations to buy bicycles?

[dasgeh]

I think CaBi should primarily be for getting around (errands, meetings, exploring) and not commuting. However, I think that CaBi should support commuters (e.g. when someone’s bike is in the shop, when someone wants to try biking for part of the commute), in part because the reality is it will be used for commuting, and because it generates good will. The problem is that if people use it for commuting, we’re generally going to need rebalancing (because people generally commute in one direction; v. getting around happens in lots of directions).

I do think that rebalancing should be focused on the primary purpose of CaBi: people getting around. So we shouldn’t be taking bikes that the “early shift” uses to get from the ‘burbs to the city back out to the ‘burbs, to enable a “late shift”, but we should be taking bikes from full docs and putting bikes at empty docs so the people who need to dash out to a meeting or to get milk can take CaBi. In the long term, people will find CaBi less reliable for regular commuting, and find another solution.

BTW, I think mstone’s point was that 0.5-1 million people TOTAL head downtown every day and it’s not realistic to expect CaBi bikes to reliably available for a non-trivial % of those every day. That would be saying that (1% of 500k = ) 5,000 bikes would be needed at the periphery every day, either by having 5,000 individual bikes, or having a fraction of that and rebalancing. Either is unrealistic. To be fair, I don’t think PotomacCyclist said anything about moving a nontrivial % of the commuting population. I think you two are talking past each other.

I think if CaBi defines the primary purpose of the service, including rebalancing to be getting around, it’s going to benefit commuters. There will still be rebalancing from the center to the periphery. But the details will be different.

And maybe some of these incentive programs would help. I find the bike-folks around here to be great with data. I imagine they have a fair amount of data on the cost of running those incentive programs v. the cost of the rebalancing without them. If the incentive programs are worth it, I imagine we’d see more of them.

[mstone]

@dasgeh 53151 wrote:

BTW, I think mstone’s point was that 0.5-1 million people TOTAL head downtown every day and it’s not realistic to expect CaBi bikes to reliably available for a non-trivial % of those every day. That would be saying that (1% of 500k = ) 5,000 bikes would be needed at the periphery every day, either by having 5,000 individual bikes, or having a fraction of that and rebalancing. Either is unrealistic. To be fair, I don’t think PotomacCyclist said anything about moving a nontrivial % of the commuting population. I think you two are talking past each other.

I’m sure that’s true. Yes, my point is that the potential pool is so large that the system has no realistic chance of dealing with any surge. If 1% of the population wakes up one morning and decides to take metro even though they’d normally use a different mode, metro won’t really notice. If 1% of the population wakes up on a fine spring day and decides to take cabi, the system collapses. (Assuming that the ability to support commuters on demand is a goal of the system.) The numbers work out the same way on a microscale. Let’s say you put a 100 bike dock in between a couple of office buildings because normal commuting demand shows that 100 people out of the 5000 people working in the adjacent buildings will typically commute by cabi. Now what happens if 1% of the remaining people decide to cabi in on a fine spring morning? You can argue the numbers and move the decimals around, but you can’t change the fundamentals. Now the response is, “well, I’m not talking about adding more people”. Well, that’s great for you, but cabi is a public, community-funded program, not a private club. Handling only existing members and assuming that either nobody else will want to join or somehow excluding them in favor of existing members is a non-starter. If you add more rush capacity, you have to assume it will be used. The idea of building peak commuting capacity for bikeshare to solve dockblocking issues in the morning is no different than the idea that adding “just one more lane” will fix the traffic problem. The only way to limit peak demand is to limit supply and let the demand adjust to reality. (And they definitely need to limit peak demand, because bikeshare doesn’t have a model that nets revenue for every trip–the cost of increasing the supply doesn’t correspond to a revenue increase.)

Let’s say you convince a couple of dozen college students to wake up early, somehow get themselves downtown, and all take bikes off the 100 bike rack to move somewhere else. You haven’t moved the decimal point in the problem, but you’ve changed the number way to the right of the decimal by a bit. And the cost of marketing to those college students, set up a program for this activity, coming up with prizes, administering the prizes, etc., is non-zero. It may be worth doing in general, if it improves overall performance; but if the starting premise is that it’s a tool to make cabi work for commuters, it’s not worth even going there.

[PotomacCyclist]

I’m not so concerned about surges. Obviously everyone understands that an extraordinary surge would overwhelm the system. I’m talking about the daily issues with dockblocking, both with full stations downtown in the morning, and full stations in areas like Columbia Heights and Capitol Hill in the evening. I read about the complaints frequently on the official Facebook page. (I consider these to be constructive criticism, just like my thread here. I am not bashing CaBi. I’m probably one of the biggest CaBi fans among the forum regulars.)

A game changer would be a couple of extra-large stations downtown, although I understand that the various aesthetic organizations (Fine Arts Commission or whatever they are) would probably block such a move. Something like a 50-dock station in Franklin Square, a 40-dock station in McPherson Square and a 40-dock station in Farragut Square. That would be a significant increase in capacity in those high-demand areas. It can also make it a little easier for the rebalancing vans to do their regular work although car traffic congestion would continue to be an issue for the vans during rush hour.

As for the cost of marketing to college students, that can be minimal. Local gov’t officials can create word-of-mouth campaigns just like they did with the previous reverse-commute program, or speak to campus cycling groups. If it doesn’t result in significant reverse-commute rides, no big deal. They don’t need to spend any money at all. Just a little time on the part of an official whose job already involves marketing or coordinating with other organizations.

Each of these efforts does not have to have an enormous impact. But the cumulative effect could be noticeable on regular days. Regardless of the marketing, rush hour commuting will always be a significant draw for potential members. The marketing message won’t change that either way. The lack of supply does convince some to switch to personal bikes, but it does lead some to simply drop their CaBi membership and use a non-active method of transportation.

I liken it to the situation on the National Mall and the National Park Service. NPS kept CaBi off of the Mall for quite a while, but that didn’t stop the demand of users from visiting the monuments and memorials on CaBi. Eventually, after NPS relented, CaBi met the demand by adding stations at key locations along the National Mall. They have already made significant efforts at meeting one of the key demands among all CaBi users, which is commuting to certain employment centers in D.C. and Arlington. They are also balancing their downtown expansion plans with broader coverage of the metro area, particularly in the current round of expansion. Again, that’s perfectly fine. Commuting is a part of CaBi (and always will be), but I understand that it is not the only part of CaBi. (With some of the isolated stations, there’s the other matter of utility and also dockblocking, because some of those stations are used so rarely and it’s more difficult for rebalancing vans to head out to those areas.)

CaBi stations are low to the ground and do not impose on the immediate surroundings. In many cases, I barely even notice the stations are there unless I’m looking for one. I point this out to address the predictable aesthetic concerns and objections from the various commissions that have veto power over decisions like bike station placement in D.C. I think a super-large station can work in Franklin Square. The D.C. gov’t already allows all of those food trucks to park next to square. Some of them are quite tacky looking. Not that I’m an aesthetic critic, but it’s strange how a row of food trucks doesn’t present an aesthetic concern but a super-large CaBi station would somehow be objectionable. Similar to the comments from the NPS about how bike stations would damage the aesthetics and atmosphere of the National Mall and yet NPS freely allowed massive, polluting tour buses to trample right through the Mall on cross streets. (Not saying that anyone here is espousing these views, just that this is likely to be an objection to stations in places like Franklin Square.)

Even if sidewalk space is an issue, the D.C. gov’t could very easily add a narrow concrete strip to seat a super-large bike station in the park. Those sidewalks and bus stop signs are already ragged and ugly. A bike station would probably be an aesthetic improvement. If this simple step were allowed, then a super-large station could easily be sited there. There would certainly be enormous demand for a large station in the square as there are numerous tall office buildings adjacent to the park. They could even have one super-large station on the south side of the park on the I St. side and another on the north side along K St.

If all the relevant parties could somehow get this approved (which I know would be very difficult), this would make an enormous difference for the daily morning rush hour. There isn’t much space to expand or add super-large stations at other downtown locations, but there is plenty of space in Franklin Square without destroying the aesthetics and peaceful nature of the park. Actually, getting more people to bike instead of driving and taking those buses (which spew enormous clouds of toxic black smoke into the square) would be a huge improvement.

These are just further ideas, ones which can help to make the overall CaBi experience better. I’m sure other people have other ideas too. I’m not looking for perfection but I don’t think they can really ignore one of the primary drivers for membership demand, which is bike commuting. Just to clarify the point about the numbers, the reason why I think it’s reasonable to cut down on the likely potential market is not just the commute distance. Casual cyclists (which make up perhaps most CaBi members) will simply not bike commute past a certain distance. A few hardy souls will, such as a lot of the forum regulars. But really, we’re talking about a miniscule percentage of total D.C.-area commuters. Most casual cyclists probably would not commute more than 3 to 5 miles on bike. Maybe less. So that knocks out most of the suburban members from the calculation. Many D.C. residents don’t drive, but there will always be a huge number that will never bike-commute on a regular basis. Others will live so close to their offices that they prefer to walk. Others will take Metro if they want to read or work on a laptop computer during their commute. (Admittedly, that can be difficult on some of the crowded trains, but not so difficult earlier in the morning.)

Just a bunch of thoughts that can show that improvements can be made as part of the ongoing and planned expansion for the system, along with non-expansion programs that can chip in for further incremental improvement. I don’t think it’s impossible to improve the system. It won’t be perfect but that’s not what anyone needs to be aiming for. Surges can’t be accommodated in every situation and that’s understandable too. But they can work on the everyday situation of the heavy morning demand. I think it’s reasonable to work toward those goals and I believe the local governments are already proceeding along those lines. But sometimes I get a little impatient. A minor character flaw on my part.

P.S. D.C. and Arlington are both seeking appropriate sponsorship deals for their parts of the network, while remaining compliant with advertising laws. (I think there are stricter state laws in Virginia about outdoor advertising.) If some of these deals work out, that could provide yet another large capital infusion that could speed up the expansion plans (assuming Bixi and Alta can work through their production delays).

Another consideration is that more and more people are moving to the District, unlike the population decline in previous decades. This has a complicated effect on CaBi. Some might live so close to work that they can walk. Others might live far enough away but not too far that biking would be the best option. And so on.

[dasgeh]

I appreciate that CaBi is very data driven. In fact, they just realized the results to a survey (which I haven’t had a chance to dive into). I’d be interested in what the driver of membership is – commuting v. getting around v. both.

Also, I believe CaBi studies where they would ultimately want stations. I could understand if the answer for DC is to have lots more smaller (20 bike) stations v. fewer larger stations (your 40 bike example) because our buildings aren’t very high, so we’re more spread out that other cities.

Finally, I imagine the biggest demand is going from/to the areas that aren’t served by MetroRail. So Franklin Square is great, but there’s a Metro Station right there. I work in Foggy Bottom along Constitution. I’ve actually converted some people to CaBi users to get downtown for lunch/errands and even around for some meetings, because the alternatives suck. There’s lots of pockets of the city like that — less than a mile from “good” areas (with restaurants, stores), but not served by MetroRail. Personally, I’d put more stations around those areas. The “suburban” stations are good for that, too, because it’s an easier sell to say “you can use it to get to lunch from the office, and it’s in your neighborhood, so you can use it to get to the grocery store, too”.

But I’m not making the decisions. And in this situation, I actually think the people making the decisions are doing a pretty good job looking at the data.

[PotomacCyclist]

CaBi gets a lot of its ideas from users and potential users, through the crowdsourcing map as well as emails and other means of communication. When the crowdsourcing map first came out, I added probably a few dozen locations on the map, all in good faith. Those were locations that I honestly felt could use a bike station. Since that time, I noticed that many of those locations now have a station. While most of those suggestions were common-sense (such as the Jefferson Memorial and Washington Monument), I think it still helps the administrators in showing just how popular and in demand some locations might be.

I had never thought of a super-large station or two super-large stations in Franklin Square until this thread. I just tossed it out there as a throwaway idea. But the more I thought about it, the better it sounded. So I wrote to Capital Bikeshare and suggested the idea in a lengthy email. I received a reply today, informing me that they would consider the possibility of super-large stations on NPS-administered locations for a future round of expansion. They also suggested that I add the suggestions to the crowdsourcing map, which I already did a couple days ago.

I think they do listen. They also track data. But for locations without existing stations, or stations that are always full or empty, the data cannot be complete or adequate. They don’t know just how much demand there is for the dockblocked stations because people can’t dock bikes at those full stations. There is anecdotal evidence from complaints and comments on Facebook. Judging just from those, there is quite a lot of demand for more stations and docks downtown (as well as in Columbia Heights). That’s probably the most common complaint I see on their Facebook page.

***
I don’t want them to focus only on downtown either. But they haven’t done that. There are plenty of stations outside of downtown and the current round of expansion focuses on expanding the system instead of filling in gaps in the core. They added a few stations in the Foggy Bottom/Constitution Ave. area in recent months. They are filling in the gaps in Upper Northwest. When Maryland adds stations to Friendship Heights, Bethesda, Medical Center, Silver Spring and Takoma Park, D.C. will add more stations near the Maryland border, to create a denser MD/DC bike network. All of this is good. But there can be expansion in more than one location, both on the periphery and in the core. They have been trying to balance the two. I have no problem with that. I’m just pointing out that the demand is always going to be very high for the centrally-located stations and neighborhoods.

[Author]

Posts