Transcript

Gina Trapani I wore my MTA badge around at home because I wanted my family to notice it. [Rich laughs] And ask me about it. Like I was just so happy to be doing this thing. [music fades in, plays alone for 15 seconds, ramps down]

Rich Ziade Paul.

Paul Ford Richard! Richard, Richard, Richard Ziade.

RZ One of the most interesting and challenging things we can do, to go from being 15, 18, 20 people to 80 people and beyond, is to give real ownership and power and control and leadership to others at Postlight. And there’s no way you reach these numbers without that being somewhat of a success, or in some cases, a great success. So I’m excited to have two of our managing partners on today’s podcast—Gina Trapani and Chris LoSacco—who have been a huge, huge part of Postlight’s success and growth over its many years, i.e. five years, but still. Excited to have you guys.

Chris LoSacco Hi there!

GT Hello!

PF Chris and Gina. So first of all, if you’ve been listening to Rich and I, we’ve been talking a lot about this, we’ve been talking a lot about taking steps back being more strategic and other people running big things. And I don’t think there is a better example of us actually getting it together and doing it—giving up authority—the MTA, which Rich, what did we do for the MTA? Well, actually, I don’t want to tell, here’s what happened. There was an RFP process. We submitted a proposal, they said we’d like you to do some work. And then it turns out that you couldn’t do that work. I couldn’t do that work. But Chris and Gina, who are here, got to do that work. Chris and Gina, what did you do?

RZ Well, Paul, I want to actually bring up something. I want to highlight the weaknesses of both Chris and Gina.

GT Oh great!

PF Oh, wonderful.

RZ I mean, this can’t just be a big love fest and just clapping for each other. No, I want to take a little bit of credit here, because this is going to be all about Chris and Gina for the rest of the podcast. But I want to take credit for one thing. This was an incredibly important project for us. We’re a New York based firm. It’s the MTA, biggest transit system in the world incredibly excited about it, and its scale was big. And we were looking to see who was going to lead the effort. And I saw something very complimentary of having Chris and Gina partner up on it. And what you have with Chris is a world class product thinker, who, literally the people in the room shrink into little marshmallows when they’re trying to make a case that he doesn’t believe in about a product direction. It’s actually great to watch. It’s a great thing.

PF Oh, it’s brutal. Let’s be clear, it applies to you and me too.

RZ It does. I’ve been a marshmallow many times in front of Chris LoSacco.

PF You cannot convince Chris LoSacco, that something is good, until he has decided it is true. [Rich & Gina laugh] And this is a very rare—

RZ It’s so true. 

PF It’s hard because you’re like ”Chris, what about—” And that doesn’t mean that Chris won’t go in there, get the work done, take one for the team, whatever. But you can’t make him believe, no matter what you do. Now, I don’t know about you. But I will believe if you can, you can actually get me like ”Paul, and if blah, blah, blah, and the business case, and blah, blah, blah.” And I’ll be like, ”Well, I guess it needs to be purple. I guess that little monster needs to be there in the corner.” You’ve reduced Chris to one component part. 

RZ Now let me do it to Gina. Now Gina’s superpower is she is a great listener, a great relationship person, someone that will build those bridges, will really hear you out really get to understand the pain you’re kind of going through. Look, by the way Gina is an engineer, Chris gets deeply involved. He’s on our sales calls is one of the best voices we can have on a sale. So I’m oversimplifying on purpose here. But boy, it was a powerful combination to have the two of them thinking about both quality and relationship at the MTA. We have done enough talking, Paul. And so I want to start this with a question. Same question for each of you, actually. How do you get going with something like this? This was a vast platform, state what it was, and then how do you get to work? Because I gotta tell you, I felt it felt good to land it but I was happy I was back in the room. Because there was a lot of unknowns, right, that were still in place.

CL Let me start by just describing what this thing was. And we have a case study live on our website so people can go read that if they want. But you know, it started with screens, digital screens were showing up across the MTA network and that meant installations in subway stations so that when you’re waiting for the train, you can see you know, messaging and advertisements, but it also meant, you know, screens on tracks for the suburban railroad stations and it meant screens on buses and eventually it will mean screens on train cars and subway cars. And so just you know, this is not just like 10 or 20 screens, this is 10s of 1000s of screens that are getting deployed across this network. And the MTA was unsure about how to deal with getting messaging to riders, you know, in a timely, efficient way across this vast network, they just didn’t deal with it before.

RZ Right. So the tools were lacking in terms of broadcasting or narrow casting to different screens across thousands of screens.

CL That’s right. And even the, the non screen channels were very dub. Like, it was like, okay, we need to post to Twitter, make sure there’s a lot for Twitter, we need to post to the website, make sure there’s a bottle website. And so literally, the interfaces that we saw were, you know, eight text boxes next to each other, eight text areas. And it was like, go nuts communications person. And one of the insights that we took from your work, Rich, early on was, let’s think about this more thoughtfully. This is not just one text box per output, it is how do we create really strong data that can be transformed to go really seamlessly into all these various places? And yes, that really helps with screens. But it also helps publishing a data feed that can be consumed by Google Maps, or the MTS website, or its trip planner, that kind of thing.

RZ As a technologist, I mean, I sort of started to sort of sketch out the high level architectural approach to this before you guys came on. So it resonates for everyone on this podcast right now. But Gina, how do you like translate that into a message that resonates for the MTA. The MTA, a huge, sprawling, public organization?

GT You know, you look at every client engagement, I like to think of it in two parts, right? There’s pre kickoff, and there’s post kickoff, pre kickoff, you’re in sell mode, you’re saying ”Hi, we’re your partner, we can help you out, we’ve got technology, and we can help you out.” And Rich drove this mostly, and I got to attend some of the meetings. But I was there in the presentation, when Rich said this is we’re gonna create pure data, and we’re gonna transform it on a pure channel basis, and it’s gonna be future proofed, and it’s going to be clean and fast, etc, you’re presenting the low fidelity vision for what you’re doing to executives, right? But then they say, ”Okay, yes, we want to do this, let’s go!” And you have a contract, and then you kick off. Now your post kickoff, and you’re in a totally different mode, which is, now we have to make this happen. And suddenly, all the details of all the things and all the people and all the needs and all the opinions and all the old systems and all the workflows that everyone’s so attached to, like bubbled to the surface, and you’re just like, there’s a moment of slight terror. When you’re like, ”Oh, we have a like, we’re on the fulfillment side now. And we’ve got to got to make this happen.” So the way that you make this happen is just—I mean, to begin with, it’s weeks, it was six weeks of conversations with getting to know everybody in the org, getting to know the nouns and the verbs. I started writing a dictionary that was like, you know, hundreds of terms just to understand—

RZ Lotta acronyms. 

GT Oh! So many acronyms, just understanding, like the landscape.

PF How did you manage that glossary? [Gina laughs]

RZ Paul, you want to know like the tool she used?

GT Yes. I mean, it was it was a text document, I go to just a flat document.

RZ Okay, so I mean, we’re glossing over this, but it’s worth noting, you didn’t have one customer and their team?

GT No, I mean, the MTA is made up of multiple agencies, right? There’s subways, which is actually two systems that the MTA bought and put together the letter, the letter trains and the number of trains. There’s buses, of which there are many kinds, local, express, select buses, and there’s Metro North, which is a railroad and there’s LIRR which is also a railroad. And then there’s bridges and tunnels, which is a whole other thing. But there’s many, many groups who all operate at different places inside the org.

CL You’ve also got different kinds of personalities within those groups. I mean, you’ve got some people who are very ready to embrace new technology and say, ”God, it’s been so painful how we’re doing it right now, I can’t wait to get to a better thing!” You know, those people who are like really engaged in those newer things. And then you’ve got people who are like, ”Listen, I’ve been doing my process for 15 years, and I know how I want it to go and you can’t—”

RZ And I’m really good at it. 

CL ”Don’t, you know, don’t take away my X or Y because I’ve learned that this is the way” even if it’s so inefficient or so problematic on the rider side, you know, they’re just these learned behaviors that just get very engraved. And so we had to, you know, early days, Gina and I and, and Jeremy Mac, J Mac, who’s our head of engineering, he joined a lot of these conversations. We had Aimee, who was one of our directors of design, like we had to understand, you know what these drivers were behind each of these groups, each of these people and ultimately let it inform the early design decisions that were getting made about about our platform.

RZ I was sort of once removed from this process as it evolved. I mean, can you really get a true sign off? Or don’t you kind of just go, like you’re showing designs, let’s say your designs are looking pretty solid, you’ve had a couple of internal iterations. And then you present them to like eight different groups? Like there is a very, very—that’s a scary moment, right? Because you could hit enormous drag at that moment, because each group comes at you with a million different—like, how did you keep moving, is the question I want to ask here. How did you keep the train—not to overuse a metaphor—but keep the train moving and on time to keep going? Because this was a project that had to ship right in the end, like we couldn’t just peter out at some point.

CL Our key stakeholder, the champion of the project, on the MTA side, she set it up such that all the key stakeholders who were involved in reviews knew that this was a v1 that was gonna ship early this year. And not all the features had to be there. And that was kind of understood. And she did a great job setting that expectation. 

PF Now, let’s talk about that for a sec, because this is the great challenge of civic technology, right? Like, getting that buy in is hard. And here are the things I’m hearing: strong internal advocate, someone who said, you know, there’s this is going to represent change. And that’s okay. It also sounds like there was a lot of sequencing, that it was like, we got to talk to 16 people, we got to talk to 16 people, no surprises. What am I missing? What you know, landing something that is going to change the way that hundreds of people work inside of an organization of this scale is always hard, inside of a government organization, even harder. What am I—what are the things that that people need to be on the lookout for? 

CL I mean, another thing I would add is don’t wait for everything to be there. Like it’s okay if you’re shipping something that is, you know, X percent of the way along and accomplishes a few key goals with the expectation that it grows in the future. A lot of projects we’ve seen that die on the line is because you’re waiting for it all to be in place before you ship and we’ve talked about this before, like you want to get something out sooner rather than later. And we definitely did that with this MTA project.

GT Yeah, I mean, in addition to the strong advocate, we also had a group of users who were hungry, and tech savvy and like ready for the next thing, right? They were in such pain and suffering that the new thing, like they were ready to try the new thing. And so we had an appetite for a better solution. And we’re also working with an organization who dig tunnels to run trains through them, which takes billions of dollars and decade’s worth of time. So us giving them anything that worked, I mean, just being able to sign in with your MTA credentials, the fact that we got that to work in, you know, 20 to 30 days with all the back and forth with IT was kind of miraculous. So actually, we were moving quite fast. And that just happened to be you know, a circumstance of this particular client.

RZ I want to highlight and something that just as an observation of seeing how you work Gina on this on this effort, every so often Gina would come in and say, ”I haven’t spoken to them in a while. I’m going to go see them.” No agenda, no meeting, which you know, has stalker like qualities to it. [Gina chuckles]

PF Once you get a badge, it’s okay.

GT I got a badge at MTA HQ, I had a cubicle with my name on it. I’ve never had a cubicle in my life. I never had a job with a cubicle. It was super cool. Listen, I was born and raised in New York City. I lived here most of my life and riding the subway in the bus since I was a kid. I was so geeked out about the fact that we even got to do this work. So, heck yeah, I went down inside of that cubicle just because I was like, yeah, here’s me, I’m at the MTA trying to make and riders lives better here in New York City where I live. Like, I’m really happy about it. So yeah, I’m not gonna lie, I had some personal feelings and investment in this particular one, even though I love all of our clients equally, and with great amounts of passion. 

RZ It’s great advice, though. I mean, forget the you know, the client vendor example here that which is explicitly what this was. But if you’ve got stakeholders, go see them, go talk to them. You get to understand sort of how they work, not just their asks, this way, every conversation isn’t framed as ”Can I talk to you for a minute?” Which is, is fine.

GT I mean, this is the thing, the conversations I had in the hallway is people just pass me by and said, ”Oh, hey!” were so much more valuable. Although I want to tell everyone to know on the show, this is the show that there was a point where Rich came to the empty HQ for a meeting and I was already in the conference room, I was wearing my MTA badge, and he looked at me and said, ”Gina, do you work at Postlight anymore?” And I said, ”No.” [Gina & Paul laugh]

RZ I remember telling you to come home more often. And I and it was such a dumb confusing thing to signal to you. Because Gina looked at me like ”What the hell man? You told me to dive into this and build—”

PF Rich was actually, I remember you were confused. You were just sort of like ”Well wait a minute, you know, she needs to be down there, but she doesn’t need to be down there.” This is the letting go.

GT It’s like two stops away on the subway. It’s not that far from our office. You could walk there [Gina laughs]

PF Well, there’s also, look, as Rich and I have grown as leaders, we’ve had to realize that when—just because I think the pandemic has helped as much as a pandemic can help—we’ve realized that if you can’t see someone, that doesn’t mean they’re not working.

RZ But it is the softer aspects of it, right, which is often not, it’s not in the agreement, we have no obligation to actually go and visit and spend time there. We’re supposed to just meet the requirements. I mean, that’s, that is the relationship side of things and relationships supply—

PF Let me make an observation here. Everyone that we’re working with in a large organization, has seen a lot of vendors come and go and seen a lot of projects succeed, and a lot of projects fail, right. And so they’re going to be real nice. And they’re going to be polite, and are going to be hopeful. But they also aren’t going to be shocked, if it doesn’t work out. So you got to actually show them that you have skin in the game as well, to bring them to bring them closer to you. And that’s just normal human behavior. I’ve been on the other side of that, you’re just like, ”Okay, let’s see what we got here.” And if what we got is somebody who is present and aware and says, ”I really want this to work” then they’re going to be, they’re going to come to more of the meetings, and they’re going to pay more attention, it’s not going to be they’re going to be really focused, and they’re going to be excited. It’s not like they weren’t going to help you anyway. It’s just like, boy, does that make it better for everybody and move things along faster if they know you’re committed.

RZ Chris, tell me about, without getting into details, the conversation that was like ”Well I also need this thing.” We never went to the contract. When you go to the contract—I’m gonna just speak now from just relationship slash legal perspective, when you’re going to the contract, it’s bad scene. It means trust is broken down to some extent, and they’re wielding the contract as a weapon. Many, many of our relationships never go to the contract. Right? And you’d never want to go there. A lot of agencies and I’m proud to say Postlight isn’t one of them, go to the contract to add work, to add ”Oh, sorry, look, this is a really smart suggestion. But it’s not what we committed to do. So we’re going to send you a change order, right.” So they go to the contract for that purpose. How did you manage the asks I’m sure asks came in that were, frankly, outside of scope. Like how did you manage that?

CL We built into the process that we wanted feedback to inform the shape of the product early on. So we, I mean, we showed our work, like all the time, in the early days. When we had wireframes and we had designs, we were presenting them and gathering feedback and letting that feedback inform what we were shipping. 

RZ The scope!

CL Yeah, the scope, exactly. We didn’t, we didn’t say we’ve already decided to need to do X, Y and Z and nothing else. Especially early days, even before wireframes, frankly, like we were playing back our interviews and saying, ”This is what we heard, we think we should focus here, does this sound right?” And we learned things even in that process. Then we also had I mean, we had a, you know, a phase two document that we were keeping track of the entire time we were working on phase one. There were several things, many things that came up as part of those early discussions where we’re like, ”Listen, we really like this direction, we have to ship software. So we can’t, you know, fully tackle this in the first version. You know, maybe we’ll do you know, a part of it in this other way. But we’re going to think about this in phase two.” And there were a lot of things that we knew we couldn’t get to but we were going to think about it the future.

RZ I mean, it’s worth noting, and this is a question for both of you. I mean, it’s it’s let’s just tell the world. I mean, the messaging that goes out to Twitter, to social media, to emails to those screens on the stations is powered by the platform that we built. Tell me about that day, when you flipped the switch. And it’s like, here we go. It’s all gonna crumble down, come crashing down. Like how is that day? I mean, and how did it go?

GT Such a great day.

CL It actually, I mean, it went pretty well! We were, we went in confident and full disclosure, we had some fallback plans in place so that if things had gone awry, and emails weren’t sending and website was broke, like we knew we could fall back to other systems. But I mean, we had battle tested the, you know, we had a beta period where we had users who were using it, and it was hooked up to fake Twitter accounts and a fake email server and all that. So we had—

RZ You simulated actual usage through—

CL Bingo! So we were reasonably secure. I mean, for the screens, we actually did in station tests where we had the screen manufacturer on the call saying, you know, ”In your operations portal, do you see the messages that we’re posting and that we’ve confirmed visually?” So we did a lot of dry runs before we actually went live, you know, in production to real riders.

PF So look, one of the things that the MTA wants, right, they come to us, we’re a smaller shop, we’re in New York City, and we’re known as being very tech forward. And you don’t come to Postlight because you want to integrate with your giant enterprise resource planning system like that’s, that’s somebody else’s job. You come to us because you want to launch something new. And so they said, you know, what stack would you use? How would you do this? Which is always great, right. And at the same time, we know that what we land has to live in their world, like this is a big IT organization, and we have to be able to give them something that will run for 10, 15 years and make sense to them. So how do you both approach that problem? Where do you start?

CL A couple thoughts. We knew, because we knew we needed to ship fast, we did use modern tech, but we met with some of those IT stakeholders early on in the process, to say, ”Hey, here’s what we’re thinking, you know, we want to use Postgres as our database. And we want to have a graph qL layer in front of it. And, you know, does this architectural architectural approach seem sound and seems like something that can live a long life here at the MTA?” And, and we got buy-in those early stages, like ”Yes, this is all, this all makes sense.” We also, we tried to integrate with, you know, their existing technology choices like where we could, right, so authentication uses their Active Directory setup, and we were hosted in their AWS account. So we weren’t choosing, you know, out of left field that this is going to be an entirely new thing that we have to introduce, for the most part, we were trying to build off of what was there.

GT Yeah, I mean, this is this is like always the balance, right? Because we want to optimize for shipping quickly, right. But we also want this thing to be a success for the MTA in the long term. We had already sold on the idea of an API driven platform, so they had already kind of they had understood, like, we’re going to produce a data feed, it’s going to go to all these different places. So they understood the difference between, you know, data, you know, client and server, that separation. So that was a good start. But then it was like, okay, you know, okay, AWS is on the table, you have an AWS account, we can host there, that makes sense. Okay, then when you choose a language, do you think what do you know what developers are likely to be at the MTA in the long term? Let’s build, you know, in that way—let’s use now we use Postgres, which has been around forever, and it’s SQL, you know, and you have to think through things like if you know, MTA IT, has to troubleshoot this thing in the middle of the night when we’re all asleep? Like, where can they look, what is logging look like etc. So you have to balance that long term stability and maintenance, you know, low effort maintenance with short term, you know, developer productivity and efficiency and top tier senior developer talent being, you know, having good velocity and being able to move forward.

PF But you know, what’s interesting here is that, basically, what we’re finding is that if people have made the move to the cloud, you can ship modern stack software. Yeah. So I mean, like, like the thing that unlocked this was, well, we’re on AWS, well, so are we, and then it turns out that everybody, everybody can kind of get going from there, which is important. Again, we’re back to civic tech, right? Like you’re bringing fast, rapid development time, for much lower cost into a very large organization, that’s a good thing to do like that, that just has a lot of merit.

RZ I mean, good credit to the MTA, also, we’ve had other experiences with clients where software was almost wielded like religion, and it was very much like us negotiating with the decisions that were made internally. They really defer to us here to a large extent, in terms of some of the more forward looking stuff that we, we leaned in on and that’s no doubt contributed to the success of the effort, like we moved really fast for a project of this scale. And they and they gave us the room, the freedom to do it. I want to close this out with a question. Or actually, if there’s advice you can share, you guys are moved on, you still pay attention to the MTA, the MTA is still a client, but there is leadership that you’ve nurtured within that effort. How are you able to do that and ship the code and meet your commitments? How did you put yourselves in a spot so that you could sort of duck into the shadows to some extent?

GT It’s kind of a wonderful seeing, you know, you rich had this vision that you sold to the MTA or not sold, but like, got buy in the MTA and brought me and Chris and on that, and there was a point where it’s like, here’s the team, okay, and now watching them have to internalize, wait, what are we doing? And why are we doing this? And why is it important? And a lot of that is showing them, you know, laying out the rationale and the story for them, but also bringing the team in to meet the users and hear the stories of what you know, what they have to do in order to get, you know, a tweet out and and the team really internalizing those those problems and issues. I mean, you know, again, this was great work. We’re all here in New York City, and this is part of our daily experience commuting to the office, but also there’s just a great deal of empathy I think on on the Postlight side, it didn’t take long for everyone, not just the senior people to really get that this was really important and interesting and challenging work that was gonna make a big difference in people’s lives. 

RZ There was strong motivation there, right, I mean, throughout. And you know, I think a piece of advice you can give to front product managers and lead designers and whatnot is, if you’re interested in getting in on the conversation, let it be known. Because you know, some people don’t like that side of it. Some people excel, and are very good at wrangling personalities and calming people down. If you’ve got that skill, it’s rarely in a product management job description, ”calm people down.” But boy, is it a skill. And if you’ve got, if you’re good with navigating people and handling that stuff, let it be known, because I assure you, the leaders and the stakeholders will leverage it. And I think, you know, for us at Postlight, we’re small, right? So you can get exposure pretty quickly, versus a very big team in a very big org where you’re trying to get noticed. And I think I think we’ve seen that really materialize for some of the leaders that have taken that project forward in that relationship forward.

CL Absolutely. 

PF It’s worth saying to this is a hard project, technically, there was a lot of work, there’s a lot to figure out. We were a little bit spoiled, in terms of our client, like not that they weren’t tough customers, but compared to a lot of very large organizations, boy, did they want us to succeed and go out of their way to help. 

RZ It’s true.

GT It’s true.

RZ We had very, really good partners there. And I mean, advice to anybody who hires an agency or goes outside for an effort, trust your partner, it’ll be increasing the likelihood of success if you do that.

PF I also cannot emphasize enough the value of the advocate internal. An agency cannot just bolt onto an organization and figure it out. You need someone in there who can, who can really sort of keep things on rails, it’s huge.

GT We also we had a unique kind of opportunity on this one to pull everyone at Postlight or at least everyone who is in the office and on this project in some way. Because we actually had digital screens that would be in the subway, on the walls in our office. And so you would go to get a cup of coffee, and one of our designers would like stop you on the way and do an impromptu, like user interview. Like, ”Did you see this notice? Like what what did you internalized? You know, do you like, do you know what you need to do next? When you see this message we did eyeline testing” and that was that was really cool. So every every one at Postlight felt like they contributed to this project in some way, even if they weren’t staffed on it. [music fades in]

RZ As a New York shop, what a great project. What a great story we’ve we’ve been able to tell with it. Credit to you guys, credit to the team that was involved. Chris and Gina, thank you so much for joining us on the Postlight Podcast.

CL Thank you!

GT Thanks for having us!

RZ Check out the case study at Postlight.com, if you need us, reach out, we’re a digital design, strategy and engineering shop. World class talent, Paul Ford, at Postlight. Hello@postlight.com is how you reach out to us and give us five—do people still say give us five stars on iTunes for the podcast? Do people do that still? Alright, listen, whatever platform you’re on, if there are stars, you can give us give us the maximum number that you can give us. If it’s a thumb, make sure the thumb is pointing upward. And if it’s a rating, if you can write a review and tell, you know, tell everyone we’re great, please go ahead and do that as well. There’s a lot of convergence happening with these platforms, Paul, so we have to speak very generically about people applauding the quality of this podcast. Alright, Chris and Gina, thank you. Have a great week, everyone!

PF Bye!

GT Bye!

CL Bye! [music ramps up, plays alone for 3 seconds, ends]