Marketing Cloud Implementation: Data Modeling, Design + Integrations, Oh My

Speaker 1: Welcome to In The Clouds podcast. In The Clouds is a Marketing Cloud podcast powered by Lev, the most influential marketing- focused Salesforce consultancy in the world. Lev is customer experienced- obsessed, and podcast hosts, Bobby Tichy and Cole Fisher, have partnered with some of the world's most well- known brands to help them master meaningful one- on- one connections with their customers. In this podcast, they'll combine strategy and deep technical expertise to share best practices, how- tos, and real- life use cases and solutions for the world's top brands using Salesforce products today.

Bobby Tichy: Hey, everybody, welcome to In The Clouds podcast. This is Bobby along with Cole and our special guest today, Stephen Rosenfeld. Excited to continue our series on implementing Marketing Cloud, and really diving in today to data modeling, design integrations. So to really kick things off first, Stephen, I'd love for you to just introduce yourself and share with everyone a little bit more about you.

Stephen Rosenfeld: Hey, thanks, Bobby. So, I am a solutions architect. I started my career on Marketing Cloud about six years ago at Salesforce. I started in the Northeast region, so I managed a lot of retailer implementations on the product. And from there ventured into integrations and after about three years, moved over to Lev. And I've been doing similar work, large implementations, cross cloud implementations, et cetera.

Bobby Tichy: And Stephen's being a little bit modest, because while he was at Salesforce, he did win the Innovation of the Year Award twice. And he's known at Lev, I would say, by everybody and by everyone I mean, me as the smartest person at Lev. So we're very thrilled to have Stephen on the podcast, that's for sure.

Cole Fisher: Now, looking at comparison of just who's on this podcast, that doesn't really say much, but Stephen is a smart dude, we'll give him that.

Bobby Tichy: He's definitely the smartest person on the podcast right now.

Cole Fisher: I don't know it Bobby.

Bobby Tichy: So as we go through this, I just a summary of what we'll be going through. First is this background of data modeling design and the strategy for it within Marketing Cloud, what that data model is based off of, what are approaches to the different integration options, we have typical integration points. And then we'll get to kind of common questions that customers and prospects have as we're talking to them, or just people that we're talking to, at conferences and things. And then also just Stephen's seen a lot. And so we'll talk through some of the things like what's the largest amount of data he seen? When does he see some lags in performance, different things of that nature. So we'll go through that as well. And so to start, I'd love to Stephens just starting to pick your brain on the background of data modeling, as we think about it within Marketing Cloud, what that data model is based off of? How we should really perceive it? And how you just typically start your discovery if you're starting a new implementation.

Stephen Rosenfeld: Sure, yeah. So I mean, there are a few key concepts that are not specific to Marketing Cloud around data modeling, right? And it's important for us to keep those things in mind when we're implementing a new account or new campaign within an account. So traditionally, we have to think about database performance, how is data getting into the account, and then margin count is built on top of SQL Server. So once we get that data into the account, it's going to be stored in a table dedicated to that account. And we have to decide, how is that table going to be used? Is this table going to be something that is used by multiple campaigns? Or are we going to pull data out of that one centralized table into campaign specific tables? And really what table structure is required by the marketing team in order to execute the campaigns that they're looking for on the technical side? So, whenever I'm setting up a new account, I traditionally simply, go through and talk to the marketing team about what campaigns they're deploying today and what their wish list has been in the past for those campaigns, right? And what has recently been implemented? What got shut down because of some data issue or otherwise really? Just to start giving me an idea of where that team is, both in their marketing campaign space and if they have technical people on their marketing team, right? Because we are we're talking about data driven marketing, right? So when we're thinking in that way, it's important for me to understand the skill sets of that other team member to find those people that are data- oriented. And really try to extract information out of them, to make sure that I'm getting things right when we talk about the structure of the account.

Cole Fisher: Yeah, that makes sense. And once we get past that initial phase in discovery and in understanding design, what are really some of the best practices when we come to think about how we perceive what the data model looks like, what best practices are around, things like segmentation capabilities, and personalization strings and things like that?

Stephen Rosenfeld: So, I mean, there are really big things, right? Like consistent naming conventions, and all that, but I won't spend any more time talking about that. I think one of the most important things is that we bring data into Marketing Cloud in a normalized format, meaning we have tables that are related, one to many, to one. And then I mentioned a moment ago about understanding how that table is going to be used within the context of the account, right? Do we have a slew of campaigns all accessing this one centralized table? Or do we need to query off of that one table and create campaign specific tables? And typically, the second is what I do. And when we are pulling that data in for a campaign specific workflow, we can denormalize it a little bit. So we start out with this normalized structure that really is very flexible and allows us to do a lot with our campaigns. But then we move more into a denormalized structure that minimizes the number of lookups that we have to do within AMPscript, for example, and that helps with the performance during inaudible.

Bobby Tichy: I think the way I've always thought about it is into your point, Stephen, we really look at what data does a marketer or does the team need to be able to send the campaigns and the automations and the journeys that they want to be able to send. Once we get all that data in there to your point, we can normalize it a little bit. So that way, if we do have some non- technical marketers, we can allow them to use the drag and drop segmentation as much as possible versus writing a SQL query since the platform's based on SQL Server. And to your point about denormalizing it. You mentioned AMPscript. And for those who aren't familiar with AMPscript, it's essentially a proprietary language. And Stephen keeps me honest, but very similar to JavaScript where you can perform more advanced functions, whether it's lookups to data or content features that maybe stretch the limits of dynamic content. And you can also call out to different services, there's a ton of different things you could do with AMPscript specifically.

Stephen Rosenfeld: Yeah, it's really close to. NET, it's a server side execution. Yeah, that gives you the ability to drive dynamic content and personalization, or just inaudible. A few of the key important notes there would be, whenever we are creating a data model, marketing law doesn't have any restrictions on how you name that table, except for special characters. And there's a few characters you can't have in there. But it's not uncommon for me to have data extensions and fields with spaces and the names. It doesn't cause like mass destruction, right? It's not the end of the world. But it does make coding that AMPscript or SQL statements when we're moving that data around, a bit more difficult to manage. So there are things like that, that we can do within the data model to make things easier for the email developers in the world.

Bobby Tichy: As you're going through these conversations and looking at the data design and talking through with the data teams, the technical teams. One thing that Cole and I have found is as we go through integration discussions, there's always the question of, what are the typical integration points? Which I think most companies already have a good idea of what they are, because they're migrating from a current service, but there's always the outlier where they're just starting to build out their marketing strategy and or what they're doing today, they don't want to continue to do. So I'm interested in one, what kind of the typical integration points that you see? And then what are some other ones that people typically don't think of? Maybe during the sales process, we outline two or three of them. But then once we get into the actual implementation, when you're involved, there's actually four or five, for example.

Stephen Rosenfeld: Well, that depends on the account executive, right? Bobby, I mean, when I get your deals, there's always like six or seven additional ones that I just never see, or that you never saw them. They're staring me in the face.

Bobby Tichy: That sounds about right. I am pretty dumb.

Stephen Rosenfeld: Well, I'm kidding of course. There's not a whole lot that on the delivery side that we would implement against that sales isn't aware of at this point, for a smaller team more unexperienced inaudible maybe. But we're typically integrating with point of sales or bringing in data from web analytics providers, and we're sending data to mainly data warehouses, right? So, Marketing Cloud is really good at storing and reusing engagement data. But it's not intended to be the long- term home of all this data, right? So typically, for that large volume, we do ETLs out of Marketing Cloud and into it. So, we've got data leaving Marketing Clouds to those EDWs to typically data warehouse, will also append things like web analytics parameters to links in the email so that Google Analytics can can pick it up. And then the API calls coming into Marketing Cloud from something like a point of sale for real time messaging.

Cole Fisher: So Stephen, when we think about... I hear a lot of really high level discussions when they talk about all the magical things that Marketing Cloud can do and in the way integrations work. And you hear something like, " Oh, well, Marketing Cloud will go and pull this out of this system and then we'll push this over here." I feel those are used a lot of times synonymously, when they obviously mean very opposite things. So can you talk a little more about what integration options there are, the different forms of integration, as well as what is really pushed and what is really pulled from a Marketing Cloud standpoint of capability.

Stephen Rosenfeld: Yeah, cool. That's a really good point. So when we think about pulling data into Marketing Cloud-

Bobby Tichy: Have I not had any good points yet? Gosh.

Cole Fisher: inaudible stream of consciousness. Come on, now I got to get back to inaudible.

Stephen Rosenfeld: That's a really good question Cole. So when we think about things like pulling data into Marketing Cloud, in traditional Salesforce and on the Sales Cloud, or Service Cloud, we can create triggers within the application that when you execute a specific task, it executes a trigger or a workflow and pulls data down or does whatever is necessary. But we don't really have that level of granularity of control within Marketing Cloud. So, when we say Marketing Cloud is going to go and pull this data in, we have a mechanism of doing that we can use script activities within an automation to make web requests outside of Marketing Cloud. And typically, that's going to initiate some process, right? I wouldn't want to do that and use that to start pulling in a massive amount of data. That's going to timeout, it's not going to work very well. But certainly, we could have something that script activity starts, it runs on an external server, maybe drops a file on the SFTP. And then we have a secondary process and Marketing Cloud that detects that file drop and pulls that data in, something like that. We would be able to use a script activity to pull data in as long as the volume's not very high. But we're talking about running this within an automation, right? So typically, the most frequent we could run that is within an hour. And it's not going to be reactive to some event, right? It's going to be on a scheduled cadence. And when we're pushing data out, typically it's the same way, right? So again, I can send small amounts of data in a predefined cadence to an external server. But the larger amount of data you try to push, the more likely you are to have that script start timing out. And so that's why I would typically end up going towards an ETL when we're talking about large amounts of data that don't necessarily need to be moved in real time.

Cole Fisher: That makes sense. And I think while we're on the page of clearing up definitions, another term that I think is sometimes thrown about, not necessarily in the most accurate way, is real time. And how do we really determine what is real time versus what is really strictly batch import?

Stephen Rosenfeld: Yeah, there's a lot of times that real time phrase is thrown around when it's not really needed real time, right? Typically, the real time interactions that we have are when someone's, doing a password reset, and we need to get that email to their inbox, but you have an end consumer sitting there waiting for this communication to be sent to them, whether it's SMS, email, et cetera. That's when we start really needing to have those real time integrations. More commonly, we have the need to be able to respond to specific data points, while a consumer's in a given workflow, right? So, I'm thinking about things like hotel reservation journeys, right communication strategies, and how do we design data to get into our data model in a way that makes sense for the timing of that journey? If my user's moving through this journey, and that user's going to hit a decision split within journey, right? It's going to evaluate my users data point on a given day, then that helps me understand how real time does that need to be? It's likely not that the moment the booking engine gets it, it needs to be over within Marketing Cloud, but it might be within an hour, two hours, right? That's, not real time but it is near real time. And most of the time, it's close enough for the needs of the marketing team.

Bobby Tichy: I think that's an interesting point you just made there of what's needed for the marketing team. I think, a lot of times marketers feel like everything should be in real time. Or I would even say the other end of it too, or maybe even a little bit more lax than others and how quickly they need data. And so I think that's typically a pretty interesting conversation of how we're doing that and the size of the data that we're removing and that sort of thing. Now, on a more serious note, the Marketing Cloud has the REST API and also the SOAP API. I'm curious to get your thoughts on-

Stephen Rosenfeld: So when we're looking at SOAP versus REST, right? I mean, there's two different ways to look at it. There's the functionality that's already there. And there's future development. So when we look at the functionality that's already there, anything that's been delivered in probably the last two to three years, I'd say is going to be a REST API. Anything that's before that is really going to be SOAP. And so there's a few different endpoints that overlap, like writing data to a data extension and triggering communications. Those two use cases are so common that both frameworks need to support them. But when we talk about being able to dynamically create things within, like an email within the platform. Since they released content builder in the last two to three years, there's a REST API for that. But when we look at something like classic content, the legacy content or email system, that was also. And so we're at a point now where we've got a hybrid of APIs that we need to be able to execute against, depending on the use case.

Bobby Tichy: And again, thinking about especially as we think about the newer features too on the REST API, you've also got the active API, the lazy API, and also the sedentary API, which is my personal favorite as well. So, glad you could share the differences there. So, Stephen, a couple of the things that you mentioned as far as all the different data elements and integrations and things like that. I'm curious, what is the... And certainly we don't want to release any client names or anything like that. But what's the largest amount of data you've seen in this single data extension table?

Stephen Rosenfeld: A single data extension was just under a billion rows.

Bobby Tichy: Okay, what was the particular reason for that? Was it just they wanted to bring in a lot of data? Or they were a very large organization that that was just their subscriber base?

Stephen Rosenfeld: It was a very large organization that had been around for decades. Their data that they were sending over was not easily manipulated on their side. So, it would take many weeks of efforts in order to get a data change put in place and so rather than creating a data model where it could potentially need adjusted in the future. We went the route of having this very tall data table that was not very wide, right? So again, we're importing this data into SQL Server and standard database performance is an issue, right? So the taller our table is going to be, the skinnier it should be. This particular table wasn't as skinny as would be preferred, right? So typically, for a table that tall, I wouldn't want more than eight or 10 fields that ours was closer to 30 or 40. But yeah, it's more than the effort that goes into changing the structure of that table in the future and knowing that it's really locked in. And once we get that data into Marketing Cloud and building workflows against that table to break down that data volume. We're not going to be doing email lookups into that table. That table is going to be subsegmented down for specific workflows.

Cole Fisher: So I assume those aren't individual records then if that's not all the contact or anything like that, unless they have a subscriber base of like one in every seven people across the globe, right? So, what was the purpose individually of that data extension? And how they functioned day- to- day without giving away exactly what they do?

Stephen Rosenfeld: Yeah, the records in this table were not one to one with-

Cole Fisher: Mm- hmm(affirmative)-

Stephen Rosenfeld: ...Right to your point. But the table held relationships to multiple brands across the enterprise. So, I as a consumer may be associated with several different brands, and I may have multiple records for each one of those brands. And because this customer needed to have both overarching and brand specific communications, we needed a decentralized table that allowed us to do that within things like audience builder. Yeah, so that was really the reason for it. It was not an individual one- to- one with a subscriber base. Yeah, not one inaudible.

Cole Fisher: One billion, that is impressive. Stephen, I want to cover off a couple of really common questions that a lot of customers or prospects, looking at Marketing Cloud have. One of which is really, aside from that discovery portion and kicking off a general Marketing Cloud implementation, what's really required from a customer's standpoint for these integrations? Everybody's curious about how much of the lifting will their internal resources be responsible for? What does that workload really look like on the customer?

Stephen Rosenfeld: Yeah, and so the response there is a question, right? It's, what do you have today? And where do you want to be? What's the purpose of our project? If you're looking to create improvements, alongside the implementation of Marketing Cloud, which most clients are, then the effort would be, we're probably going to have to create some new ETL processes on the client- side to send that data to Marketing Cloud for your marketing team to execute against. If you already have ETLs built and you're sending them to a specific SFTP, Marketing Cloud can use that same file, right? We created a new data schema for every client that we implement. So, in that way, Marketing Cloud is very flexible, and it can fit to your current model. But the question is, should it? And so when we talk to the marketing team about their campaigns, and how they're running today? What's their wish list? And what's been a roadblock for them in the past? And why are they moving to Marketing Cloud? That's where I start to be able to answer that question of, " Well, how much lift is this going to take?" And if there are concerns around that on the business side, then we start to put a roadmap together, right? Say, " Okay, well, this piece of work was going to be a lot for your IT team and they don't have that bandwidth right now. But this section of data over here, you're already got. So, we can bring this into Marketing Cloud and maybe execute different workflows against it, then you're able to get in your current ESP to unlock some new features. But we'll do that while your IT team prioritizes this other batch of work." Does that answer your question Cole?

Cole Fisher: Yeah, it sure does. Another really common question that comes up is, is there a limit really to the frequency or the volumes that Marketing Cloud's are able to store and just data?

Stephen Rosenfeld: Yes, I mean, every database is going to have a limit to how much it can hold before you start having performance issues, right? So, whether it's the number of tables that are in your account, or the size of a specific table. So yeah, there are things that we've hit in the past around this, but traditionally, we can work around them. When we think about the limit of data coming into Marketing Cloud from something like an API, there's no hard limit. Marketing Cloud does have limits on how much volume they want you to bring in from the REST API and after you exceed some threshold, you'll sound some alarm bells. But within Marketing Cloud, I've never had an instance where they actually shut down the API and don't allow execution against it. And like I was saying, really it's more how you manage the data model itself, right? I've had clients that create three new data extensions, every time they want to deploy a campaign, they deploy let's say 100 campaigns a week. And then two years later, everything slows down, because there's so many different pieces of the application that just load a list of data extensions. And when you have that many data extensions, that the application is trying to fetch and display, it starts to slow things down. So, there are different pieces of the application you have to consider when you're doing your data model, right? So certainly, I wouldn't want to do a send against a table that has half a billion rows in it, unless I absolutely had to. And in that scenario, I would have to take other precautions to make sure that send's performance shouldn't be doing dozens of lookups, against other tables, things like that.

Bobby Tichy: When we think about integration specifically, are there certain products or platforms that Marketing Cloud does not integrate with?

Stephen Rosenfeld: That's a fair question. I can't think of a product that I've been asked to integrate with, that I've not been able to. The reason being, Marketing Cloud enables the ability to perform HTTP requests from either scripts that are scheduled or within emails at the time of deployment. And so, any other system that has a REST endpoint or not even an API, it could just be an open endpoint, we can pull data in and store that data in a table within Marketing Cloud and start executing against it, right? So, I don't want to say that there's no platform out there we could not integrate with but certainly I feel that if there is, it's likely an issue with the other platform, and not something that's due to Marketing Cloud architecture.

Bobby Tichy: It's been pretty neat to see the amount of integrations. And when I say integrations, I mean, productize integrations that have come out, especially in recent years and primarily thinking of ones like form assembly and segment and these historical tools that maybe weren't central to the Marketing Cloud Data Model. But as you know, the Salesforce acquisition of ExactTarget five years ago becomes more and more... Six years ago I think, actually becomes more and more embedded. And these customers see that the Marketing Cloud tentacles are getting further and further. It's been pretty neat to see those. I think the top one, obviously, that comes to mind for me is the integration with Sales and Service Cloud. It's interesting too, because I think there is some stigma out there, that Marketing Cloud has to be integrated with Sales and Service Cloud. And obviously, the three of us know that certainly isn't the case. But as we think about folks who actually utilize Sales and Service Cloud and integration to Marketing Cloud, the best practice from the product and technical standpoint, is that we integrate or we push data into sales and service cloud first, because the integration between that and Marketing Cloud, the connector is based off of the lead or the contact ID and the Salesforce side. But one question that we often get is, can leads go into Marketing Cloud first, or leads or subscribers or however you think of it, MQL versus SQL, whatever that might look like. So I'm curious, in your experience, Stephen, where you've seen that work well, and just some of the technical elements that allow you to do it?

Stephen Rosenfeld: Yeah, so you're absolutely right. Anytime we have Marketing Cloud being implemented, it's ideal to have a CRM in front of it, right? The last thing you want to do is try to manage your consumers within your marketing platform. It's better for us to be able to feed that data back and have a reliable CRM in front of it. When we think about managing those leads to your question, we talked a lot about how we can expand Marketing Cloud's data model to fit the needs of something that's pre existing or supports marketing teams, but I didn't mention as much that the data model that's inherent behind Marketing Cloud, right? So this goes back to the dreaded list conversation, right? So lists are what Marketing Cloud was originally developed on top of, but it didn't scale the same way that data extensions do. But at the end of the day when, at the time of send, the list model is still executed against. And if you implement that list model in the right way, you can absolutely bring your leads into Marketing Cloud. And what gets a little messy is when you have leads coming in from multiple sources, and they don't share the same identifier, then it becomes a conversation around what makes sense for us to use as an identifier for your leads. In most of the time, it's going to be email address, which is going to be an inherent identifier, right? That's the mechanism that we're going to be communicating with them through, lets use it to identify them. And that allows us to consolidate those and scrub out duplicates easily. When we talked about those leads converting into contacts, we want to make sure that we continue to respect their opt in preferences. And if we implement the list model in the right way, we take advantage of functionality within Marketing Cloud to suppress users in efficient ways, then we can ensure that even though they've converted to a contact, let's say they're using Service Cloud or Sales Cloud, and we're now using that contact ID as their subscriber key that of how Marketing Cloud identifies that individual consumer, we can still ensure that they're not going to get an email from Marketing Cloud. Because when they were a lead, they opted out of marketing communications. So there isn't a baseline data model that we're building off of. And if we take advantage of that baseline data model the right way, we can cover those types of use cases.

Bobby Tichy: I think that we were actually just on a call earlier today with a company that's evaluating Marketing Cloud and they have CRM. And they were asking how to go about this. And so the two main ways that I know how to go about doing this primarily journey builder, and then also through a script activity if we want to go that route. And I think there are a couple other ancillary ways too. But one thing they were asking us was, let's say, I go to an event and I have someone's email address and Marketing Cloud, but I get all this other data about them. And I want to upload that into Marketing Cloud, and then have that carry over into Sales or Service Cloud. I think there's a real conversation of where are we importing data? And why? So why would we import that data into Marketing Cloud versus putting it directly into Sales and Service Cloud? But also, to your point, Stephen, what is the identifier that we're using? And really the matching up capability between those two is only really as good as the raw data that we're getting. If I leave in CRM with one email address, but I leave in Marketing Cloud with another email address, and there's no other identifier that can match me together. Then, obviously, it's not going to be possible. But as long as we've got a identifier that allows us to match between them, everything hopefully should work just fine.

Stephen Rosenfeld: Yeah, and like I was saying, it's always best for us to have that CRM funnel in front of Marketing Cloud. To your point, if I'm bringing data into Marketing Cloud, and then moving it to something like Service Cloud, then the reference to my individual consumer is not the Service Cloud ID, right? So, even if those emails did match at first and I did have that established connection, if I change my email address, then it breaks that, right? So, if we bring that data into Service Cloud, then we can call workflows and rules around managing that data and we gain the benefits of that, the CRM in front of it and then everything naturally flows down into Marketing Cloud. We typically try not to push things up to the CRM in that way.

Bobby Tichy: For sure, well, thanks Stephen a ton for your time. I really appreciate diving deep into the data side of things and the Marketing Cloud side. Just one thing before we get to completely unrelated, with connections 2019 coming up in Chicago this year, want to let everyone know that Lev will be there, I will certainly be at the Expo Hall. And then we're also having a happy hour event as well on that Tuesday night, Tuesday, June 18th. You can visit the Lev website and sign up or visit any of our social media areas as well. Also, once again, if you guys have questions or thoughts, anything like that, that you'd like to share with us, you can write us at intheclouds @ levdigital. com. Again that's intheclouds @ levdigital. com. So let's move on to completely unrelated and Cole I'm gonna let you start.

Cole Fisher: We've done a fair share of traveling this week and I was talking to a customer on the phone. So, I had a meeting yesterday from the airport where I was trying to hide in a hallway of the airport and I'm hunched over my laptop like a grizzly mother sheltering her cub, trying to filter out any of the noise and I did a really poor job of it too. Because they come on and gate changes and announcements and things like that over the back. It's not really good for the acoustics of your standard Zoom or GoToMeeting or whatever. So, I got to the meeting, and then I was talking, a follow- up today with the customer and they were asking, " Did your travel go all right?" I was thinking I actually it was the second time, I was in LaGuardia at the time. It was the second time leaving New York in the last six or eight months, that I'd actually left on time and not had delay. inaudible coming in wasn't lucky but leaving New York I just had really bad luck. And so I'm actually curious, Stephen and Bobby what you guys' most hellacious travel stories have been?

Bobby Tichy: Oh gosh. Well, before I go there, I will say I wish that all airports had two things to help pass the time. I wish they had those phone booths that we've seen in offices where you can just go in there if you've got a call. They're not completely soundproof, but they're at least sound resistant. So, you get away from everything. I always wish that they had those and I would even pay for it. If I had to pay five bucks for half an hour or something like that. I would be more than willing to do that. I also wish they had just walking treadmills or something. I hate just sitting at the gate and just waiting. It just look like most the time I'll just start like pacing between terminals or something or between gates because I just get really antsy. And my most hellacious travel story though. Oh, man-

Cole Fisher: Bobby, before you get there, I will suggest this. If you're looking for like a walking treadmill or anything like that, just go up the down escalator. I mean, you might get some stairs and it might be a little inconvenient for everyone else but it's a great way to burn calories.

Bobby Tichy: I really like that. I might do that the next time. If they ask me what I'm doing, I'll just tell them that you recommended this approach.

Cole Fisher: No, I just said that you eat lots of paint chips. And don't ask questions.

Bobby Tichy: Wait, hold on paint chips not a part of everyone's nutritional diet?

Cole Fisher: As long as they're lead base. They're cool.

Bobby Tichy: Okay, I just wanted to make sure. I can't really think of most why... I'll flip it on you. So this wasn't an awful travel experience. This is actually a really good travel experience. My wife and I were on our way home from our honeymoon. And I said at inaudible. We both had slept a couple days off and so we had some time and they were offering like an $ 800 Delta credit. And you got to fly first class if you bumped flights and took a later flight. And it was the first time I've ever flown first class. And first of all, I didn't know they give you free alcohol or free snacks. And also the first time for those of you who have ridden first class, you get this little warm towel. I had no idea what to do with this warm towel. I didn't know. I should have waited to see what other people did with it. But I just started washing my face with it and apparently that's not what you're supposed to do. You're just supposed to wash your hands with it. But that was a great experience, especially, this is a probably, how long ago did we get married? Hopefully my wife doesn't listen to this. Almost seven years ago, so like, getting 800 bucks plus flying first class was pretty incredible.

Cole Fisher: If that's the most hellacious travel story you've come across, I really don't feel bad for you whatsoever-

Bobby Tichy: I'm sure I have slept in an airport at some point-

Cole Fisher: Well, this hearkens back to when I was... I think it was in Newark, where you and I both had flights out at almost the same time. We both boarded our flights, I think yours was like 20 minutes ahead of mine. Both boarded our flights. Yours was like the last one to take off. Mine was the first one not to take off. And then they ended up shutting everything down at all the Newark airports because of a rainstorm-

Bobby Tichy: No snow it was-

Cole Fisher: Yeah.

Bobby Tichy: Yeah, I could see because it had started... It was like a blizzard. As we were on the runway and I was thinking there's no way I'm going to get off this thing. And then you guys ended up having to stay like another 24 hours, right? Like you didn't even get off anytime soon.

Cole Fisher: No, You got out just by the skin of your teeth. And we sat there on the plane before the plane, for another hour and a half. And they delayed and delayed for a few more times. And then we sat in the airport for another nine hours before they're like, " Yeah..." This is on Friday, they're," Yeah, it's canceled. And we're really not going to have any flights until Sunday." I was like, " Oh, great." So I think crosstalk-

Bobby Tichy: ... Thisjust goes to show that the universe likes me a little bit better than they like you, Cole.

Cole Fisher: Apparently. All right inaudible I'm going to challenge you here. So mine was probably what? 2016, 2017 was my first time flying to Vegas, the trip went really well. And then on the way back, my flight left, I think it was probably 6: 30 ish. And Southwest did not let us check in until about, or we couldn't get through security until about four hours before the flight I think it was. We actually got there quite early, and couldn't check in or couldn't go through yet. And so, we went and sat down. And sitting there with my wife and time comes that we can finally go check in and we get up and we walk back to the kiosks. And there's a line from every kiosk out the door and around the airport. So this was when Southwest went down. And it was just insane because we were there early and it was super quiet. And then we had to wait, so we found this quiet corner of the airport, we come back and there's just lines everywhere. So, then we go through... Luckily we were checking in with four hours before our flight. So okay, so we went through that line, our flight gets delayed. It was supposed to take off at about 6: 00, 6:30, ended up getting delayed till 11: 30. I think we got moved a couple times in there. They're canceling flights left and right. But then they don't cancel ours, they actually canceled the flight to Denver and gave us their plane, so we had to deal with all these people around us that just had their flight canceled and they're handing that flight off to us instead. And we get on the plane, the flight attendants make their speech. And then we sit there for probably half an hour and then eventually they get on there and they say everybody's got to get off the plane. The crew's been in the cockpit for too long today. By the time they get to Indi they're going to be past their inaudible limit. And so that crew was not allowed to fly anymore. So we had to get off the plane. Everyone's luggage was taken off the plane and Southwest had... We were in the airport in Vegas and there's just luggage everywhere. And they tell us, they're not going to bring that luggage out because there's too much luggage out, they're not gonna be able to manage it. So they told us to just leave and go. They'll get a hold of us for the luggage. And my wife and I looked at each other like we don't have anywhere to go, let's just stay here. Actually no, we booked a hotel room and we booked it for the wrong night because it was after midnight. And so we ended up staying there but then right after everybody left then our luggage comes out. And so it was so chaotic, getting home but we ended up getting home the next day. It was probably 2: 00 PM.

Bobby Tichy: The great part about all this is that I was with you Cole in your most hellacious and I got up scot- free. I was also with Stephen in Vegas when he had this experience in mind and my wife and I got out of Vegas on Southwest before they went down without any issue.

Cole Fisher: You are the worst. I can hear you smiling and it makes me very inaudible slash your tires the next time we hang out.

Bobby Tichy: I feel like I should knock on wood and do all of these things because I feel like I continuously just get lucky.

Cole Fisher: I just don't feel sorry for anything about you, Bobby.

Bobby Tichy: Well, thanks a lot guys for the time and for the episode. Stay tuned for the next episode of In The Clouds.

Cole Fisher: And we'll catch you guys later. Thanks.

Stephen Rosenfeld: Thanks, guys.

Bobby Tichy: Yeah.