The Washington, DC region is great >> and it can be greater.


Innovation resistance at Metro, part 2: The Google bugaboo

Yesterday, I discussed the way the burdens from a "top-down" approach hinders innovation far more than any potential benefits to Metro.

Photo by the author.

That addresses the first of the two major issues, releasing data to developers. Metro did release its schedule data, but under a restrictive license, and when asked about doing the same for bus position data, Board member Gordon Linton expressed skepticism, largely out of fears that people are out there "lining their pockets." He, and Metro staff, often seem to be thinking about Google when they say this.

After all, Google has a lot of money, right? If Metro's going to do something, shouldn't they get a piece of that?

It'd be great for Metro to get some money if they could. I fully support Metro getting as much revenue as it can, from Google or any other company. (See the disclaimer about my previous relationship with Google at the bottom of this post.)

However, the important question is not how rich the company is that Metro is dealing with, but what the actual market value is of the resource being negotiated. Alstom, which manufactured the most recent railcars, is an enormous company as well. When they manufacture cars, they make a lot of profit. Is that unfair? Maybe they should be paying Metro for the right to make some cars, instead of the other way around?

Metro pays Alstom because railcars have a value determined by the market. There's also a market price for transit data: zero. As Michael has pointed out in the past, every other major U.S. transit agency, and numerous others around the country and the world, have all given the data away for free. They believed that free was a reasonable price.

Gordon Linton noted in his rebuttal to my testimony that these others haven't asked for revenue. That may be true, and IP considerations aside, it's okay for Metro to ask. They could also ask Alstom to give them rail cars for free. But in both cases, that's not going to happen. So far, there's been no deal after years of delay. Whatever Metro is asking for, they apparently aren't going to get it. And if Google paid Metro, then everyone else would start asking for money, too. Quite simply, it's clearly better for Google not to work with WMATA than to change the existing, settled market price for transit data.

It's also important to keep in mind that Google is probably not making any money off this service. Linton, Metro staff, and others seem to assume that this must have a value because a big, rich company is asking for it. That fundamentally misunderstands the way Google is structured.

Google acts much more like a startup than a big company in many ways. The typical company will only release a product if it's worth a considerable amount of money. After all, any product requires development resources, marketing time, and more. There's a roadmap, and everyone works on the top priorities on that roadmap. If you launch a product, you better be willing to invest millions to market it to make it successful, otherwise it's not worth it. Google, however, tries very hard to maintain a startup culture where people can just whip out projects because they feel like it.

In an industry like shampoo, where there are manufacturing costs and huge marketing costs, that's how businesses work. In technology, it doesn't need to be that way because it's pretty easy to make a halfway decent Web site. And Google has worked hard to maintain some of that nimbleness. If an engineer, or a group of engineers, wants to build something, like adding transit directions to Google Maps, they can go ahead and do it.

There's still a coordination cost, as a product manager has to get executive buy-in, the user interface has to be immaculate, and lawyers have to sign off if copyright could be even remotely involved, but you can launch a side product at Google with only about 10 people involved. This is why Google often gets mocked for having so many "beta" products, many of which don't go much of anywhere.

It's easier to think of Google as one extremely profitable business plus a very large number of very small, completely unprofitable technology startups that are all funded by one VC firm or philanthropic foundation. The profitable business is search and content advertising. Google makes just about all its money from the ads on search and the content ads on millions of other Web pages. Then it's got a lot of other stuff, most of which just loses money. At least when I was there, Gmail, for example, cost way more in capital costs (all the computers that store all of your email) than it made in ad revenue. It's valuable strategically, but a big money-loser. The same applied to Maps. YouTube was a ginormous money sink. And so on.

I don't know if Google is making any money off ads on Google Transit, but either way it's a rounding error. Maybe the ads make enough to cover the engineering and legal time, plus the computing resources. Probably not. They can't be making much. This product started because some engineers were disappointed that Google Maps let you find out how to drive and not how to take the train. The company is supporting this product because people like it inside and outside the company, not because it is a major strategic focus or even a minor one.

It is, however, valuable to riders. Even if the trip planner were perfect, which it's not, many people are already looking up businesses on Google Maps and would find it much easier to get directions with one click. Many people are using iPhones or Android phones and want transit directions on their existing map application. Others don't know about the Metro trip planner.

Metro staff, and Linton, are so obsessed with the imagined revenue that they won't get. Meanwhile, they're missing the bigger issues that affect riders. Next, we'll put this whole issue in perspective.

Disclosure statement: I worked for Google from 2001 to 2007. I did not work on Google Maps or Transit. I am not coordinating with them on this series. I do still own some Google stock, but am selling it in fixed increments each month without regard to its performance. I don't believe that Google Transit influences the price of the stock by even a penny.

David Alpert is the founder of Greater Greater Washington and its board president. He worked as a Product Manager for Google for six years and has lived in the Boston, San Francisco, and New York metro areas in addition to Washington, DC. He now lives with his wife and two children in Dupont Circle. 


Add a comment »

Google, however, tries very hard to maintain a startup culture where people can just whip out projects because they feel like it.
"Excuse me while I whip this out..."

by anonymous on Sep 30, 2009 11:21 am • linkreport

There is some credit to metroÂ’s debate. Google is getting revenue off the ads that run in the background on GoogleÂ’s server when I search Google maps for mass transit. Also if apple sells a app that includes the dc metro maps and schedules then I believe dc in entitled to some % of that money.

by matt on Sep 30, 2009 11:44 am • linkreport

The overall problem is that if WMATA asks for the money they'll likely get nothing because it's easy for Google or an App developer to say "no thanks, there are other transit agencies out there".

Just because revenue exists doesn't mean you're entitled to some.

by Michael Perkins on Sep 30, 2009 11:53 am • linkreport

Penny wise, rider foolish.

by Jasper on Sep 30, 2009 11:57 am • linkreport

Is there honestly anything stopping some interested transit nerd from taking the publically available Metro transit data and putting it into the format needed for Google Transit?

by jfruh on Sep 30, 2009 12:03 pm • linkreport

@jfruh: It's already in the correct format. The only thing stopping Google Transit is Google and Metro not signing an agreement.

by Michael Perkins on Sep 30, 2009 12:07 pm • linkreport

@Michael: If it's merely an agreement, why haven't we seen someone try to broker such an agreement (by someone I mean someone not you :)? If we can broker mideast peace, we can try to have someone broker such a deal with Google Transit.


by Jason on Sep 30, 2009 12:36 pm • linkreport

@Jason: When did peace in the middle east break out? I missed that headline.

by Michael Perkins on Sep 30, 2009 12:40 pm • linkreport

This article is just plain wrong. The analogy to Alcom sucks, the proposed value of the data is a big hand wave, and claiming Maps ad views and, more importantly, search history, aren't important to Google is just plain silly.

I love your site, an I think you're generally insightful, but this article isn't at al perceptive.

by jcm on Sep 30, 2009 12:45 pm • linkreport

This article is just plain wrong. The analogy to Alcom sucks, the proposed value of the data is a big hand wave, and claiming Maps ad views and, more importantly, search history, aren't important to Google is just plain silly.

I love your site, an I think you're generally insightful, but this article isn't at all perceptive.

by jcm on Sep 30, 2009 12:45 pm • linkreport

Wow, I double posted, typoed, and wrote Alcom instead of Alstom. I guess you really got me fired up! Sorry about that.

by jcm on Sep 30, 2009 12:46 pm • linkreport

Does Rand McNally pay EZPass for listing which roads are toll roads on its maps?

Does AAA pay the managers of tourist attractions, rest stops, or hotels before including hours and locations in their TripTiks?

This whole debate is ridiculous. Shame on WMATA. Save a buck now. Make two later.

by SDJ on Sep 30, 2009 12:52 pm • linkreport

Furthermore, WMATA ought to start charging for attending its press conferences, and for access to its press releases. After all, the Post, WJLA, etc., etc. might might be able to charge a little more ads based on Metro's intellectual property. That seems to be the argument, that "their" data puts additional eyeballs on Google's ads.

by Paul on Sep 30, 2009 1:06 pm • linkreport

So what is Metro justifying their reluctance on here? Is it:
1) Data is valuable because it just is - it's our IP, and we're not letting it go without remuneration.
2) Licensing our data to other entities on which riders might rely takes directing those riders out of our hands; should bad directions discourage ridership, it should be _our_ bad directions.

My guess is that Metro's argument is closer to the latter, but this post seems to assume more of the former. Neither actually makes much sense to me (the latter slightly more), but this post could come off as a strawman argument.

by Lucre on Sep 30, 2009 1:24 pm • linkreport

Lucre: They haven't made the latter argument. They have made the former one, among others. See the Zimmerman-Linton exchanges in the related posts.

jcm: I welcome disagreement, but it might be more persuasive if you explained why you don't find this convincing, as opposed to just asserting that it's not.

by David Alpert on Sep 30, 2009 1:28 pm • linkreport

Didn't WMATA make the latter argument with the beta version of NextBus?

As Lucre notes, neither argument is particularly compelling, but I swear I remember discussing the accuracy of NextBus's beta site here...

by Alex B. on Sep 30, 2009 1:33 pm • linkreport

Also, Michael, what I, and I think jfruh are wondering is, given that the data is out there, available, and more or less ubiquitous except perhaps for the specific format, which Metro won't license to Google, to what extent can that data be considered public domain, or a kid in a basement reformatting it be fair use? Essentially, what's preventing what jfruh said from happening is a fear of getting [successfully] sued, right? What I want to know is how justified is that fear? I don't mean to blindly assume you're an IP lawyer, but maybe someone who reads this would have an informed opinion.

by Lucre on Sep 30, 2009 1:35 pm • linkreport

You're right, David, I should be more specific, and I apologize.

1) The Alstom analogy
This is in no way applicable. Alstom's business model is to build and sell rail cars. Google's is to sell ads. The analogy would be more apt if, say, Metro were attempting to hire a Google to write a new trip planner on Metro's site for them. Instead, we have a situation where Google wants Metro's data in order to drive more traffic to Google's site.

2) The market price for Metro's data
You assert the market price for Metro's data is free. This is wrong for a number of reasons. Google is not the only company that may be interested in this data. MS and Yahoo both have mapping sites, as well, and there's a quite a few other types of sites that may have interest in this data.

There's also the possibility that metro could do best by improving their own trip planner, and generate revenue of their own site. They've already done the hard part(the routing engine). Any decent developer could improve the interface. Hell, you could do it with google's api and get a portion of the map's ad revenue.

Just because other transit agencies have given their data to google free does not mean that metro need follow suit.

We both know that the data has value. No one else can supply the data. I think it would be foolish to give a valuable asset away without fully understanding its worth.

3) Maps, and particularly transit's value to google.
Neither one of use knows specifically how much value google gains from maps and transit. You assert, though, that it's negligible, and imply that the project is mostly altruistic. I say that's wrong.

Firstly, Google display ads on the page, which are likely generating some revenue. Secondly, google gains mindshare, which, as far as I can tell, drives lots of their products. Google wants to be the one stop shop for finding information on the internet. Thy want to be the first place people go when looking for maps/directions Lastly, and possibly most importantly, google wants to serve us better ads. There's a reason they give away analytics. The more data they have about our interests on the internet, the better. Knowing that I was searching for a transit route to safeway is valuable to google.

Google has devoted a ton of money to maps. They and MS are spending a fortune taking satellite photos, licensing navteq data, taking street view photos, etc. They can't think they are throwing that money away.

4) Your comment that Google or an App developer can say "no thanks, there are other transit agencies out there".

Well, sure they can. However, it's in google's best interest to have metro's data available, and metro is the only source for that data. Will google pay? I have no idea. I know they'd prefer not to. They've written the interface, the routing engine, and the data format specification. More data streams are free money for them. But google's best interest and metro's best interest aren't necessarily the same. If google won't pay, will ms? Will Yahoo? And if they do, will google follow suit?

I don't pretend to know what the best solution for metro is. It may in fact be that the benefits of just publishing the data to all outweigh the lost revenue. But I don't believe anyone, including metro, knows the answer to that question yet. I think it would be foolish to for metro to make a decision on this without better understanding the market.

by jcm on Sep 30, 2009 2:20 pm • linkreport


The data is available for download, but only under license from WMATA - that is, it's NOT public domain in any sense. It's still very much copyrighted and legally you are only allowed to use it under the terms WMATA specifies.

Google could already technically go and download the data under the currently offered license agreement just as anyone else could. The problem is that, IIRC, one of the terms in that agreement reserves to WMATA the right to charge you for it whenever it gets the whim. That term seems specifically aimed at Google. (And of course that's a PITA threat for independent developers to have hanging over them too.)

by jack lecou on Sep 30, 2009 2:27 pm • linkreport


The fundamental error I think you're making is thinking about WMATA's best interest.

What WMATA's focus should be is the rider's best interest. And I fail to see how keeping this data sequestered and harder to access is good for riders.

by Alex B. on Sep 30, 2009 2:40 pm • linkreport

I agree, Alex B, that WMATA's job is to move people as efficiently and pleasantly as possible. I'm just not convinced that the best way to accomplish that goal is to give the data away. If it is, so be it, but I think it's important they examine all the alternatives.

I think we can both agree that the best solution would be one which generates revenue *and* gives robust tools to riders. If that's a possibility, then it's worth exploring.

by jcm on Sep 30, 2009 2:56 pm • linkreport

When thinking about the revenue impact, it's good to put it in perspective. Last year WMATA mentioned that a nickel fare increase would result in an extra $30M in revenues. We're talking about less than $1M in revenues per year here, so that's equivalent to a fare increase of a little less than two tenths of a penny per ride. If you ride a lot (1000 rides per year) that would add up to roughly $1.60, or one ride. So WMATA is willing to fight this one out in the best case to save its riders $1.60 per year in fare increases.

And that's IF this is worth a million a year, AND you ride a lot.

by Michael Perkins on Sep 30, 2009 2:59 pm • linkreport


1) I'm not really seeing why the Alstom/Google analogy is dis-analogous. Not saying they aren't, but I'm not getting your argument. As far as the argument here is concerned, all that matters is they both buy and sell products, at market prices.

2) I think you're not really acknowledging the market price argument here. You say "I don't believe anyone, including metro, knows the answer to that question yet." But that's not true. We do know, under any reasonable standard: Nobody is selling this data right now. Period. And WMATA is literally the last major hold out. Neither Google nor anyone else looks like they'd be willing to buy it anytime soon. (And don't shortchange the can of worms it would open if Google or Yahoo or anyone DID agree to pay one agency or another - they'd have to do it for everyone. They're not likely to do that.)

So how many years are we supposed to wait for WMATA to greedily sit on the data, inconveniencing riders, before we can finally conclude there's no price for it?

by jack lecou on Sep 30, 2009 3:08 pm • linkreport

@jack lecou: That's what I'm worried about. Metro will continue with this consultant study, it will say that the data is worth "$X", and then Metro will feel like they have to hold on to the data even if every time they ask for "$X", everyone says "No, thanks".

I've corresponded with certain iPhone app developers, they have told me that if the data required a license fee, they would not bother making their app support that agency. I suspect that Google would make the same call. I have not been able to get them to state that on the record.

by Michael Perkins on Sep 30, 2009 3:20 pm • linkreport

Yeah. Shorter WMATA: "What's wrong with you people? My $500,000 consultant says it's a perfectly good refrigerator. And a nice compact size for an igloo, too. So I'm just going to sit down here in the snow until someone buys it!"

by jack lecou on Sep 30, 2009 4:22 pm • linkreport


1) Alstrom Analogy
While it may not be perfect, I think the Alstrom analogy has some validity. Alstrom makes a physical product that, ultimately, MetroÂ’s customers use. Google makes a software product that, ultimately, MetroÂ’s customers [would] use. Google is essentially offering to provide a service to Metro at little-to-no cost to Metro. Would Google get something of value from the deal, in addition to payment from Metro (which, in this case, is zero)? Sure, theyÂ’d get some small amount of ad revenue (probably less than the cost of development and hosting) and, probably more importantly, mind-share. But, Alstrom gets something of value from Metro too, besides payment for the cars. They get the ability to name WMATA as a client to potential customers, and the knowledge and designs from building Metro cars. Should Alstrom pay Metro some type of referral fee every time Alstrom gets a new customer partially based on their work with WMATA?

2) Value of Metro Data
I completely agree that Metro data has value. But, just because something has value doesnÂ’t mean the market price for it is non-zero. Several large jurisdictions offer this sort of data for free. IÂ’m sure theyÂ’d love to get money for it, but they really want third-party tools available for their customers. Ideally they want both, but they didnÂ’t think that would happen (now evidenced by Google and WMATA).

The question isn’t so much about how much Metro data is “worth”, it’s about how much someone would be willing to pay for that data. I think those are slightly different, because other things factor into a company’s decision on how much to pay for something. Odds are pretty good that someone interested in making money off of WMATA data would also be interested in making money off of NYCT data. For now, people can get NYCT data for free. Really, I think the only data transferring hands is being given away, so the current market price is zero. But, if someone starts buying data from WMATA, other transit authorities are going to want money too. So the total cost of buying WMATA data is much, much higher than what would be paid to WMATA.

Could the market price change in the future? Sure it could. But, at least for now, the market price for WMATA data is zero. And currently thereÂ’s no reason to think anyone has an incentive to buy this data.

It makes sense for Metro Directors to think about what theyÂ’re potentially giving up by giving away the data, but I think what theyÂ’re doing is pretty ridiculous. TheyÂ’re spending $500k to figure out if figure out how valuable something is that other transit authorities have apparently already decided isnÂ’t particularly valuable. Even WMATAÂ’s current data suggests it isnÂ’t terribly valuable. They make about $68k in ad revenue from the web site. At this rate, itÂ’s going to take them more than 7 years of ad sales just to make that back. ThereÂ’s no way thatÂ’s a cost-effective way of making a decision on this.

by Andy R on Sep 30, 2009 5:11 pm • linkreport

We're talking about less than $1M in revenues per year here, so that's equivalent to a fare increase of a little less than two tenths of a penny per ride.

I myself would be willing to suffer a fare increase of as much as a full penny per ride just to end this nonsense and get some transit goodness on my phone already. How about it WMATA?

by jack lecou on Sep 30, 2009 5:43 pm • linkreport

@jack lecou: I'll second that. Heck, I'd even give them a full nickle.

by James on Sep 30, 2009 8:02 pm • linkreport

Can someone please forward this thread, namely Andy R's words, to the Post, Times, Examiner, WTOP, WRC, WTTG, WJLA, and WUSA? Pretty please?

It's nearly the 2010's, is muckraking extinct?

by Jason on Oct 1, 2009 11:44 am • linkreport

@Jason: I've pointed this story and idea out to Kytja Weir at the examiner. I'll also note that you could forward this thread yourself to everyone on your list. I'm sure they have a media tips line.

by Michael Perkins on Oct 1, 2009 11:54 am • linkreport

I would add a #3 to Andy's points: The value of Google's service.

I think that neglect is one thing I find particularly frustrating and illogical about this "debate". It's true that Google recoups some of their expenses with ads, maybe even comes out slightly ahead. But Google isn't just sucking up the data and magically "profiting" off of it. They're doing it by providing a genuinely convenient, valuable service to WMATA's users. I'm not seeing the justification for WMATA to extract a pound of flesh from that equation, even if they could.

Anyway, if it were all about who's profiting from the transaction, we could probably make a good argument that WMATA should be paying Google for the valuable privilege of being featured in their map directions: there's no question Google's got a lot of users, many of them tourists, business visitors, residents with Blackberries and iPhones needing to get to unfamiliar places. So I think everyone agrees Google has the potential to generate a lot of extra off-peak trips. How many? I don't know. But unless my math is badly wrong, even just a few tenths of a percent increase in trips would be worth millions in fares.

by jack lecou on Oct 1, 2009 1:09 pm • linkreport

@jack lecou: I sent a comment to Metro's LunchTalk forum for tomorrow's discussion expressing this. I don't take Metro off peak nearly as often as I would simply because it's not convenient to figure out how to get where I want to go. A couple of great examples are that Metro’s Trip Planer doesn’t integrate with the GPS on my Blackberry and doesn’t understand locations that aren’t stops/stations. This means that I to know which Metro stop/station I want to use at both ends. Google Maps solves both of these problems. Trip Planer is also extremely frustrating to use when it doesn’t recognize “Chinatown” as a destination, or when it gets spaces between words wrong. For example, I’ve had it suggest that I meant to enter “Shadygrove” when I had entered “Shady Grove”. Upon correcting my “mistake” it then suggested that “Shady Grove” was the correct spelling, all the while not giving me any useful information. Google Maps just does all of these things better.

by James on Oct 1, 2009 3:14 pm • linkreport

Back of the envelope, it looks like if Google inspired, say, 200,000 regular WMATA riders to take just an average of two extra off-peak trips each month, that would work out to more than a 1% increase in trips. Which would something like a full order of magnitude more revenue (in fares) than even the most optimistic estimates of what the data is worth in ad revenue.

by jack lecou on Oct 1, 2009 3:32 pm • linkreport

Metro General Manager Catoe is answering questions during an online chat tomorrow at noon. You can find out how to submit questions ahead of time using the link below. This might be a good opportunity to raise this issue.


by Andy R on Oct 1, 2009 7:09 pm • linkreport

I also submitted a question. Not something he wants to talk about, evidently.

by jack lecou on Oct 2, 2009 1:35 pm • linkreport

@jack lecou: My questions were all ignored as well, as expected. Same as the previous months.

by James on Oct 2, 2009 4:29 pm • linkreport

Add a Comment

Name: (will be displayed on the comments page)

Email: (must be your real address, but will be kept private)

URL: (optional, will be displayed)

You can use some HTML, like <blockquote>quoting another comment</blockquote>, <i>italics</i>, and <a href="http://url_here">hyperlinks</a>. More here.

Your comment:

By submitting a comment, you agree to abide by our comment policy.
Notify me of followup comments via email. (You can also subscribe without commenting.)
Save my name and email address on this computer so I don't have to enter it next time, and so I don't have to answer the anti-spam map challenge question in the future.


Support Us