Curious customers, cool tools, and open data

Our metrics site was originally designed to provide an interactive version of the information published in our MTA board books (like from our team’s presentation at the March 2026 meeting!).

But we know all too well that our board members aren’t the only folks who care about MTA data—far from it. In fact, we often frame our discussions around a set of eight user stories to guide the features that we build and data we publish.

Take Jamaica on-time performance for the Long Island Rail Road as an example. LIRR policymakers may care a lot about the systemwide average for this number. Tech educators may want to give a data lesson by showing how LIRR delay data can be “rolled up” to figure out which trains are the best performers—and worst offenders. Journalists may want to see how a bridge rehabilitation or snowstorm affects MTA service patterns.

One website for all these different use-cases? It sounds nice, but it’s also clear our existing metrics site has pain points for everyone these days. 

We on the Open Data team are committed to ensuring that the state portal isn’t the only place people can use MTA data. After all, the external metrics site has proved indispensable to many of our stakeholders—college students, good-government advocates, Albany policymakers, and more. 

So as we refine and rebuild our external data tools, we’re thinking carefully about that line between what our non-technical users need to understand our data, versus what we might know our more technical stakeholders can do themselves in Excel, a PowerBI or their favorite GIS software.

After all, why should they? Most New Yorkers aren’t completing a subway challenge—they’re commuting to work and school, visiting their families or running errands. Bus wait assessment is an important metric for our team and bus operators to pay attention to, but folks told us they care much more about how frequently their bus runs, during rush hour and near their stop, than how frequently buses run in general.

For example, a group of CUNY students used our data on Automated Camera Enforcement to demonstrate how the program would speed up their commutes to and from school. Subway Stories used our ridership data to illustrate the patterns of movement and life in the city. And journalists at Bloomberg used our archived realtime data to help their readers understand how often LIRR and Metro-North commutes are disrupted.

In the journalism world, there’s an old adage that all news is local. The same can be true for our data—so as we build out new tools, we’re going to be especially focused on making things that help our riders explore transit in the places they care about most.

Though the MTA was created by the State of New York in 1965 to unify transportation services in the NYC area, transit service is still influenced by many other government agencies in ways that might not be obvious to the average New Yorker.

Here are a few non-exhaustive examples:

• Bus stop and bus lane infrastructure, used by MTA buses, are built and maintained by the NYC Department of Transportation, which is not a part of the MTA.
  • However, the MTA sets the location for those bus stops in conjunction with DOT and other agencies, like in the recently-completed Queens Bus Network Redesign.
• Most land around NYC subway tracks is not owned by the MTA, but by New York City itself.
• JFK and LaGuardia airports are owned by the City, but leased to the Port Authority of New York and New Jersey, so programs like the free Q70-SBS must be coordinated closely with both organizations.

It’s complicated. And that complexity is mirrored in how complex it can be to give a good, data-informed answer to a tough data question. Sometimes, an answer that’s accurate or actionable might require combining, cleaning and inspecting several different datasets that are owned by multiple different agencies.

Ultimately, when our users have a question about transportation, they expect the MTA to have the answer, and we’d like to be able to give it to them—even if it’s not found in our data. As such, we’re really interested in publishing tools that incorporate data from partner agencies like NYCDOT, Port Authority, and more. 

Of course, we know our MTA data best. But we’re always excited to keep collaborating with our colleagues at other organizations to make sure that MTA stakeholders can get good, relevant answers to their questions.

We on the MTA Open Data team were amazed with just how much folks loved the maps we shared with them—not just as ways to play with data, but also as vital visual aids.

For instance, we had a visitor in the middle of our time at the Oculus begin discussing issues she had with the M9 bus. As we spoke, it felt like we were talking past each other, but pulling up the beta version of BusTime, and showing her the new map feature changed the dynamic of the conversation immediately.

“Yes! Oh my gosh,” she said. “Can you zoom in for me?”

As we rushed over to the computer, she put her finger on the screen.

“There! Now my son goes to school there,” and moving her finger an inch, she said, “and that’s where we normally get on.”

Maps are great ways to give people a common language. They represent the world—especially corners of it as complex and messy as New York—in ways that are quickly parsable and intrinsically linked to our physical environment. Like our friend on the M9, the right visualization can unlock so much information that wouldn’t be possible otherwise. So we’re thrilled to look at just how we can use our mappable data to enable more experiences like that.

At the pop-up, we got about 40 requests for new datasets. Some of those data we do have and publish on a regular basis, some we have and are in the process of publishing. But some data we just don’t have.

For instance, one of our most common requests from the pop-up was for datasets on crowding, both in stations and on subway cars. LIRR and Metro-North both have handy diagrams that demonstrate how busy each car is, helping weary commuters nab a solo seat.

Why don’t we have this for the subway? The short answer is that most cars on our railroads have technology that weighs the cars as they roll towards their destination, and this weight serves as a proxy to show how crowded each car is. Subway cars don’t have the same technology. There could be other ways to publish that data someday, but for now, it just doesn’t exist.

There are a lot of stories like that, so moving forward, we’re going to be particularly interested in how we connect folks with the best data we do have, as well as gathering and publishing new datasets that can help users explore more of their questions.