The Data Stack Show - 252: What the Heck is Happening in Data Right Now with the Cynical Data Guy

Episode Date: July 9, 2025

This week on The Data Stack Show, Eric and John welcome back Matt Kelliher-Gibson for another edition of the Cynical Data Guy. The group explores the current state of data engineering and team dynamics while critically examining the evolving landscape of analytics engineering, dissecting the hype around the modern data stack and its tools. The conversation also explores the challenges of data team management, including headcount reductions, rising technology costs, and the struggle to maintain efficiency. Key discussions revolve around the need for open standards, the impact of AI on data roles, the complex hiring practices in tech startups, and so much more.  Highlights from this week’s conversation include:The Evolution of Analytics Engineer Roles (1:53)Job Titles and Role Consolidation in Data (3:20)Standardization and Open Data Standards (7:51)SQL as a Universal Standard & Vendor Lock-In (11:58)Modern Data Stack: Hype vs. Reality (13:29)The State of Data Teams in 2025 (18:12)Morale and Job Market Realities for Data Professionals (25:17)Bonus Round: Extreme Work Culture Satire (28:41)Honesty in Hiring and Team Building (33:18)Challenges of Building and Leading Data Teams (37:31)Final Thoughts and Takeaways (41:15)The Data Stack Show is a weekly podcast powered by RudderStack, customer data infrastructure that enables you to deliver real-time customer event data everywhere it’s needed to power smarter decisions and better customer experiences. Each week, we’ll talk to data engineers, analysts, and data scientists about their experience around building and maintaining data infrastructure, delivering data and data products, and driving better outcomes across their businesses with data.RudderStack helps businesses make the most out of their customer data while ensuring data privacy and security. To learn more about RudderStack visit rudderstack.com. 

Hi, I'm Eric Dotz. And I'm John Wessel. Welcome to the Data Stack Show. The Data Stack Show is a podcast where we talk about the technical, business, and human challenges involved in data work. Join our casual conversations with innovators and data professionals to learn about new data technologies and how data teams are run at top companies. How to Create a Data Team with RutterSack Before we dig into today's episode,

we want to give a huge thanks to our presenting sponsor, RutterSack. They give us the equipment and time to do this show week in, week out, and provide you the valuable content. RutterSack provides customer data infrastructure and is used by the world's most innovative companies to collect, transform, and deliver their event data wherever it's needed, all

in real time. You can learn more at ruddersack.com. Welcome back to our favorite- You jinxed yourself. Oh my goodness. I jinxed myself. They should leave that in. They should leave that in.

They should leave the whole intro. The whole intro. Yes. Welcome back to our favorite monthly installment of the Data Stack Show, where we go deep into the bowels of corporate data America with the Cynical Data Guy. Matt, always a pleasure to have you. Hello. Thanks for having me back, I guess. Is that a new mic? It's not a new mic.

I'm just normally in studio, so I don't have to use this mic. All of you people went shit setting everywhere. Yeah, for all the listeners at home, we normally do this in person. We've got, we're phoning in from three different time zones, and it's 8 o'clock, I think, where you are, Eric. It's 11 o'clock. Three time zones, two continents. Yeah, several continents.

Okay. We need to get to it because we have a really fun bonus round one if we have enough time. So let's start out talking about analytics engineers. This is a nice post. I will read it. And then, cynical data guy, you get first take on this one. Analytics engineers have always existed before DBT, before the modern data stack, before cloud warehousing. These things do not define the role of an analytics engineer.

Were they always called analytics engineers? No. the role of an analytics engineer. Were they always called analytics engineers? No, they were called data architects, data engineers, data modelers, and even data analysts. These people were always building sustainable data practices with best practices at the forefront. They just didn't have a way of defining what they were really doing beyond their responsibilities. Now we finally have a name for these data roles and it's not going anywhere. Is it? Is it not going anywhere? I mean, it existed before the name existed, so it could exist if it got absorbed into some other titles. I mean, I don't know. That feels like something

that's there, especially as data teams condense. It feels like an easy one to go away on some data teams and just absorb into other roles. Also, I don't know how many of them were making sustainable best practices based off of what I've seen, but yeah, they were making stuff. Agreeable data guy. I don't know. I think I want Matt to keep going on this one. Yeah. Agreeable data guy. I don't know. I think I want Matt to keep going on this one. Yeah, the funny thing to me about the post is like we actually completely left out, and I guess it depends on when you start your time horizon, but we completely left out database administrators.

I mean, that's like the title, the historic title from the people that are doing this. That and like system administrators, it was essentially some combination of those two. But yeah, as far as roles, like, I, it's not going anywhere from a sense like, I think the, like, and that like, yeah, the work is gonna continue. I would agree with that. I don't, but yeah, I mean, saying that the job title, especially with all the changes that we're seeing, saying the job title is still gonna be analytics engineer, that seems unlikely. Especially, I mean, that, when did that job title, especially with all the changes that we're seeing, saying the job title is still going to be analytics engineer, that seems unlikely. Especially, I mean, when did that job title start cropping up like the last couple years? Right. I think dbt was the original source. Yeah. So I like it in the sense that like, yeah, I do think we're going to start working more,

continue to work more in this way of version control and some better practices. But I don't know. I don't know if I'd be quite so bullish on the job title, not changing. It feels like it will change and it will have the word AI in it somewhere. Well, I was just saying, that's what, there's a lot of these ones that like,

it's the same role, but every once in a while, we have to jazz up the title to try to get more money for ourselves and things like that. So I mean, I could definitely see that happening, especially if it becomes, if the perception is that the role is too narrow, let's say, in what it does, because I've seen some of that criticism before,

that it's this in-between role that doesn't do enough. It's like, well, we'll give it a fancy new name and we'll stick AI in there and then we'll get more money for it. Well, what I do see happening, and I don't see how this, I don't see how we get away from this, I really think we're gonna have the consolidation of roles. Because when you get productivity gains, like that's what people do. You have, especially bigger orgs, like you have less distinct roles and more like general roles and more general roles.

So instead of dividing data between engineers and analysts and architects, there's probably going to be less in a lot of orgs, less differentiation between those roles. And same for, I think, a lot of technical roles. I agree. I think HR is going to play catch up because at a small company, no one cares, right? You just say, I am responsible for this. But at large companies, you have leveling and career progression and all of that. And so of course, you have to have some clear definition. And that matters, that's important.

But for sure, what is someone who is a skilled user experience designer and front-end engineer and can wield all these AI tools? What's their job title going to be? I don't know. I mean they can do, you know, and increasingly, you know, there's the ability to dip into the full stack realm as well, right? Which those people used to be complete unicorns but are becoming more and more common. Even before AI that was happening. Yeah. You've uncovered it's going to really screw up HDR's attempts to do a form of legalized collusion on salary setting there. More on that later.

More on that later. Yes, more on that later. This is a post from Kirill Bobrov. I'm sorry Kirill if I pronounce your name incorrectly, but he's a senior data engineer at Spotify. He wrote a post that got a huge amount of traction. So I'm just going to read his post here from LinkedIn. Data engineering now with 30% more bullshit. That's the title of one of my latest posts.

And apparently it hit a nerve. 40,000 reads across medium, substack, and X, dozens of DMs saying finally someone said it, but this isn't just a rant. Here's the real noise I'm seeing. Data fabric sold as magic plumbing until every connector still breaks and metadata needs manual upkeep.

Medallion architecture repackaged old-school multi-layer warehousing with shiny new names and brittle DAGs. Zero ETL buzz that quietly shifts the complexity of cleansing, schema drift, and retries elsewhere. Modern data stack, just another marketing badge slapped onto generic over-engineered tool combos. What happened to delivering real value over these buzzwords?

I want to hear your war stories, the chaos, the duct tape fixes, the real hacks you had to invent because the modern stack failed you. If you're in the trenches building something that actually works, drop it below. Let's raise the bar together. What was the not rant part of that? Maybe it's in the article, but I don't see it. This isn't just a rant. Just a rant. OK, that's, you know.

But I mean, it's like, you mean a lot of this is just marketing? No. Really? I mean, yeah, we I think that it's one of the conclusions that a lot of us made, especially kind of like the post, let's say, like 20, 21 period was. There was a lot of. There is a, 21 period was, there was a lot of marketing to try to sell a lot of tools with very narrow slices of things. You know, because how do you make money? Anything you couple and then you decouple.

And that was the decoupling stage. Yep. Man, my take on this is something I've been thinking a lot about is essentially I think I can speak for most people in the data realm. I think we all want common open standards and commercializing that is just like hard, if nearly impossible. Like for example, like I think just about every data person, if you could tell them, hey, we reinvented SFTP, FTP, like, here's this new thing.

Or we, or, or we, like, and it's just as ubiquitous as FTP is slash was. Same thing for filemores. Like, we have CSV, but it's this better thing. Now, there's attempts at that. So like Parquet is an attempt at that. Iceberg on top of parquet. Like we're trying on the file format side and we're trying on the a little bit on the storage side but it's kind of locked in right like it's s3 or it's you know Google or Azure storage. So that one's less open right like you're actually using something that's

proprietary like s3. It's cheap but it is still proprietary. So in my mind like I think for Utopia for a lot of data people it's something like S3. It's cheap, but it is still proprietary. So in my mind, like, I think for Utopia, for a lot of data people, it's something like that, where you can do all this like neat stuff or zero copy of this or that, but the commercialization path is tricky. And I think Iceberg has seen that. And then we're seeing, you know, like Duck Lake came out, we'll probably see more of that people, you know,

in that space, but that's what I see is that would actually be different, but it's ironic that we already have that in a sense that like any company in the world could can create a, you know, SFTP server, any company in the world can read a CSV file. We just want better versions of those two things. And the commercialization path there is just not, like, it's not obvious. It's like, how do you, like, develop the quote world standard and, like, raise a bunch of money and, you know, all the things, right? So I think it's tough. I mean, you're also dealing with the fact that every time you come up with, like, well, hey, we could do this and it would be really great and it would be an open standard and everybody would be able to use it.

And then everyone goes, that's amazing. You know, it would be better if it just worked on my platform. Yeah. And that's, what's interesting to me is like, like for example, with what, in the AI realm, so you've got like MCP, right? In the AI realm, like the, which open AI and, you know, Anthropic both have kind of adopted that protocol. So to me, if you wanted the new version of, I mean, I guess Parquet is probably as close

if we're talking CSV versus a file format. But if you want the new version of that for data, Iceberg is the closest thing, I think. Databricks supports it. Snowflake kind of, like they're all like a little bit supported and then you get to the catalog layer and it's a mess and everybody has their own thing. So it's just not, it's not a first-class like good experience yet.

And if it doesn't get there, then it will never be ubiquitous and that's the challenge. And I think part of that is like, cause if you think of something that it's not storage, but it's something that every data person uses that kind of has that, it's like SQL, right? Like SQL is a universal standard. Yep.

Everyone has their own kind of flavor where they, you know, geek around the edges, but there is a universal standard to that. But I think for something like that, you need something like an ANSI, right? Where you're like, Hey, this is the outside thing of we're saying an organized standard, it must fit these things. If you go outside this, you're no longer in this, you're not compatible. And that could potentially force some of that

where it's like, sure, you can have your iceberg version, your Databricks version, but it still has to do these core things. And that seems like the most likely thing to win is like pick a iceberg as a standard. And then if you use iceberg with Databricks, it does some cool things that doesn't work for snowflake but there's at least a core like fully functional thing that works across everywhere. Right. Similar syntax like that.

like switching companies and you can be like modern, you can be like fully immersed in like one set of tools on quote a modern data stack and then move companies and have to relearn every single tool. Sure, Python, SQL, things like that are gonna be the same. But it's like, man, I gotta learn Databricks when I know everything in Snowflake and then I have to relearn like some ETL tool. Like it's painful if you move companies every few years,

like a lot of people do, like it's the learning curve, it's no joke. relearn some ETL tool. It's painful. Two quick thoughts on this before we move to the next one. One is that I agree with so much of what Kirill's saying. Generally I would say, yeah, the medallion architecture and the marketing was way oversold, the promise relative to the reality. But I will also say that Spotify's data problems are probably different than a lot of companies.

Because they are extremely sophisticated, they have been for a long time It was actually just moving away from a system where the tools and the database were the same thing. And you were just abstracting pipelines out like from the data store, right? It was just saying, okay, well, let's like separate concerns here like at a very fundamental level. I mean, funnily, yeah, go ahead. Well, I was just going to say, it's kind of funny how the term

was just co-opted by so many of these super specific tools Yeah, go ahead. you know, in Informatica. If you take the SQL Server stack, that's essentially like modern data stack, but you have different vendors. If you take your reporting layer that used to be with SQL Server, your integration layer, your database layer, that's still the stack. At the core, then there's catalogs and all these modeling or whatever, but it's not that different than what you were doing. You were just typically all, all within Oracle or SQL server, whatever, right. Or it was just integrated. Yeah. And it started as something where like it made sense because a lot of these

places, they did some things well, they did other things not as well. And it was like, man, this would be really better if I could mix and match parts of them. But then we just sprint to the extreme as fast as we can to where you lose all the utility of it. Right. Yep. We're going to take a quick break from the episode to talk about our sponsor, Rutter Stack. Now, I could say a bunch of nice things as if I found a fancy new tool. But John has been implementing RutterStack for over half a decade. John, you work with customer event data every day and you know how hard it can be to make sure that data is clean and then to stream it everywhere it needs to go.

Yeah, Eric, as you know how messy it can get. years and going. Yes, I can confirm that. And one of the reasons we picked RutterStack was that it does not store the data and we can live stream data to our downstream tools. One of the things about the implementation that has been so common over all the years and with so many RutterStack customers is that it wasn't a wholesale replacement of your stack. It fit right into your existing tool set. Yeah, and even with technical tools, Eric, things like Kafka or PubSub, but you don't have to have all that complicated customer data infrastructure. Well, if you need to stream clean customer data to your entire stack, including your data infrastructure tools, head over to rudderstack.com to learn more. Okay, that was a good one, Kyril.

Thank you for that. We'd love to have you on the show, actually, to hear about Spotify. So if this gets around to you, please jump on the show. We can dig in and you can tell us what we got wrong in assessing your posts. Okay. I'm just going to read all of this, actually. like AI budget today. Vendors needed revenue, they started increasing prices or marketing org. The team is now two to three people, and the data tech spend is on track to grow

from 500K to 750K to 1 million, even though the data team is doing less than they were before. They just don't have the time or bandwidth or expertise to cut tools or migrate. Bigger vendors need growth at all costs with the hopes of IPOing and showing investors something that resembles liquidity. As tech spend for data teams increases and headcount shrinks, migrate.

new data role because demand is low given the market environment. That's just my take. One day soon, it'll shatter. I've been starting everyone. John, do you want to start with this? Well, this is yours, Matt. Matt found this as nobody will be surprised. Yeah, I'll start with this one. Yeah.

I think, I don't know. I don't know. The head count I think is real. I do think the headcount reductions is real the 500 to 750 to a million and Spend I don't know what the where that goes unless the company. I mean all this stuff is Usage based so I guess I guess the idea here is there's fewer people So they just keep spending money on tools to try to bail them out

I guess the idea here is there's fewer people, so they just keep spending money on tools to try to bail them out. No, I think it's, no, a big part of it is just literally raising prices when you're under contract. It is a really interesting point. It's kind of threaded in there, but I think what the author is referring to is a company will just, if you have someone under contract

and they implemented your infrastructure, right? Especially if it's like fairly core infrastructure is a company will just, you're not going to rip that out. your customers like pay you for that, right? 30% premium on literally your existing stack because they know you're not going to turn And they just raise prices. Yeah Well, especially because it's so usage based and it was like you built all of these things to Track and store and pick up all this data when it was small and cheap and nobody was really looking at it And then it starts snowballing and suddenly they go. Hey Why is why are my cloud bills going up 50% a year?

And then they raise prices on you in the middle of that. And you're like, whoa. Yeah, I think that's how you get that type of creep on. Putting my executive hat on, I've never had a vendor do that to me, but you can't just like take the price increases. Like you have to push back, cause you have to remember sure are you would it be a pain to migrate? Do you not have the talent to migrate? Yeah, maybe but like they don't necessarily know that and they can't you lose logos like you're there just as much in the spot

We're like they're not trying to lose logos either. So that that's true But I that's true and I was extreme saying 20% but 10% happened That's yeah, totally and if you're continuing to just ingest data Especially on data pipeline tools. You can easily double and triple bills quickly and that's less about price increases and more about Consumption I think but honestly the solution to this is not migrating typically It's assessing what people actually use and killing stuff that's not being used. I mean, most companies, if you adjusted that, you could cut most of your bills in half, I believe. And I think part of it is,

because I've seen this in real life, is where stuff got set up. It was inefficient when it was set up, but nobody cared because it didn't cost that much or it was there and, oh, we're gonna make a fortune off of this or whatever the thought process was. And then as you go through these rounds of like, they cut people and people leave and things like that, you now get into a situation where it's like, hey, we're, you know, ingesting

and dealing with all of this data over here and it's going up, you know, 20% a month. And the people who are there are like, I have no idea. I don't have the time. And that would take us like six months to just try to untangle it and figure out what's going on there. And we're just being like in that, just ad hoc request,

bearing down on us constantly. And then if you start doing stuff like cutting the team, leads, like the heads and stuff like that, who's the person who's doing the hardball negotiator? You, nobody. Yeah. Well, I mean, this actually happened. I actually have like worked with a client hands-on with us. They were about to, they were not spending a ton. They were spending a ton, but didn't need to be. They're spending about 250,000 a year more than they needed to. And the root cause was like, one, as you might imagine things nobody was using that just need to be turned off. But that takes a lot of

digging and technical expertise to do. But the biggest one was, I think Matt, you essentially just said this, the team that made the stuff that was consumption based couldn't see billing. They had no idea how much it cost. And that's not atypical. A lot of these platforms are set up to where there's decoupled and it's compute based. And the team, if they just had visibility into what they were doing, could probably make better decisions. But they don't even have visibility even if they wanted to. Well, and that I think is a two part one of the tools don't always give you visibility

into it or they make it harder for you to find. But then also a lot of data teams for whatever reason, leadership, whether it's like the data leadership or whatever, they're like, we don't share budgets, we don't share. They don't know. So they're just building. What do you think about the morale question? The morale statement?

It was like, morale is super low because of all this, that people stay in their jobs because of the market. I think that's real. Or the morale statement. It's like, you know, morale is super low because of all this, but people stay in their jobs because of the market. I think that's very real. I know a number of people who have been looking for, in the data engineering world, have been looking for jobs, one of them for a couple years, actually. But again, hasn't found a perfect fit, has stayed because it's like, well, like this is a fine job. You know, I get paid pretty well. I don't love it. But you know, I haven't found something that I like better and I think that's a typical story. Right now.

I think also if you were changing jobs, especially in that like 2021 time period when like salaries really were going like haywire one time period when like salaries really were going like haywire, you could, I think that skewed some people's expectations where they're like, well, I don't want to go down to whatever. And it's like, well, no, you were just overpaid for a year or two, you know? And I saw that with people that, you know, we had people who left on my team and they would come and say, you know, they're gonna get 40,000 more than what we were paying them and it was like, well, you can go ahead and take it, I'm never gonna tell you to not do that.

But I was like, so they're going like, they're gonna be in for a rude shock in a year or two because they're not gonna make that much money again, at least not in the near term once things kind of corrected themselves. Yeah. Okay, that's a good lead in for the bonus round. Anything else on that one? There was a lot in that post.

One other quick thing I'm curious, Matt, what's your take on the AI thing? Because I do think people are like, oh, yeah, I'll just jump on the AI train and try to get a role, kind of a data AI role. I think, well, I think part of that is like, did you learn anything from this last cycle? You know what I mean? Sure. And so it's like, if you were think part of that is like, did you learn anything from this last cycle? You know what I mean?

And so it's like, if you were in here and it was like, hey, times were good and we were building stuff and you know, you didn't pay attention to the costs and you weren't, you didn't ask for what was going on. And you know, maybe you didn't, maybe you were kind of caught in like, well, we're building this foundation. And then it found out that like, oh, nobody was really using it. And that was part of it. It's like, okay, you're going to go to AI.

Are you going to learn from this or are you just going to run the same thing again? Because I think that will be a problem if you try to do that. You know, if you run the same kind of playbook, you're going to end up in the same position, which is a lot of money spent and suddenly everyone's not happy with you. Even more money. Yeah. Yeah. Yeah. And just keep a tally of how much money you've spent for companies over the last five years. I do. Yeah. I think it's interesting. It seems like there is a general subtext of FOMO.

If I'm not figuring out the pleading edge, am I going to be relevant? It's a general subtext of FOMO. If I'm not figuring out the pleading edge, am I going to be relevant? However, I think a lot of people would generally agree it's so early. Right? Right. Right. I mean, companies have started and been put out of business. I mean, MCP technology itself is changing dramatically week by week. and then put out a business. I mean, MCP technology itself is changing dramatically week by week. What a time. I mean, it really is a fascinating time.

It's not surprising that there's a lot of unrest, I think, out there in the market. Okay. Bonus round? Bonus round. This one's great. Okay. This is Austin Nasso. Sorry if I mispronounce your name. Nasso. He's a founder, an AI futurist, an innovator, and pioneer, 100X mindset. Austin, we'd love to have you on the show.

Yeah. This is a great post. Recently, I started telling candidates in the first interview that our company, Raild, Austin, we'd love to have you on the show. It's a high stress and extremely competitive environment with zero tolerance for low output. We also pay below market rate salaries and do not offer equity. They also have to come to the office every single day. However, we do provide all of the equipment for them for their jobs. It's provided on an interest accruing loan. Sorry, I tried to make it through. We deduct the loan payments from their salary and the remainder net taxes hits their bank

account. There is often little to no remainder. At first I felt bad about telling this to candidates, but I realize it's better than them getting surprised when they join. Does anyone else do this with their company? Where do you find your indentured servants in 2025? Austin, well done. That is great work. Yeah, I mean, this is just the raw version of an extreme version of what you see some companies try to pull. They're like, you know, we're it's all about we're like a family environment, a.k.a. you must live at the office.

Yeah. A.K.A. you will see us more than your family. And we will pay you like you're the 12 year old working in our shop so it's very funny, but yeah, I do like the I Felt bad at first, but it's better. They realize it up front and then be surprised. Yeah. Yeah I think I could see that going poorly being surprised on that Yeah, I think it's a really tricky I think it's a really tricky, I think it's a really tricky thing. And I think it's getting trickier just because the startup environment,

especially as it relates to funding has changed so dramatically, you know, the bubble sort of burst and it's just a very different environment than it was, you know, five years ago. And I think what's really difficult is, you know, balancing a great company ago. that's large enough to support your thesis, it's all about execution, right? So that's tricky. Part of the reason that people dig in so hard on startups is because especially if you're there's some sort of outcome,

But it is, it's really, it's a, I love the post because it's really provocative because it's not, there isn't really a straightforward answer. John, you're chuckling. I want to, this is so funny because I'm just looking into his, it's like, what company does, is he actually hiring for? So it's one of the things he's involved in, Austin's involved in, we'd love to have you on the show is the comedian's roast techies. And he says, okay, so get this, tech is everywhere and it's enabling people all over the world to do more revolutionizing modern science

at an incredible pace and changing the way we live our everyday lives. Tech also enables grown men and women to ride scooters to work. Finally, there's a show just as socially inept as you are. Just good stuff. I do think one thing with hiring, though, that's hard beyond this is trying to balance that idea of to a

certain extent, you are trying to sell, right? Come to our company. Here's all the great things that you that we're trying to do. But on the other hand, I think at least from my point of view, I always wanted to be more kind of honest so that you kind of knew like, you know, yes, I'm here and I want to be here but like, here are the things that are not always great about working here and whatever

they were. And HR does not always like it when you say that to people because they want you to do the whole raw cheerleader thing most of the time. Whereas, I thought it would be, I always felt it was important to just kind of be like, you know, here are the things that you'll probably find frustrating, right? Here are the quirk things and here, but here's why I'm here. Like, and a lot of times I feel like as a hiring manager, what you're selling people on is not necessarily like the company is this. It's more like, here's why I'm here doing this. Yeah. There was a who did I can't remember. I can't remember who, we can look it up and

put it in the show notes, but someone talked about the anti-hiring email where they get a candidate, they're excited about the candidate, the candidate's really excited and they would send this email, say, hey, I just want to let you know, like every job, there's good days and bad days. Here's a description of what a bad day looks like in this role. And they would literally send an email, right? And it's like, well, if they were like, okay, like, I'm still in, then that was a good sign, you know, as a way to screen them. Here's, I think, the weird part about this is I think this is contextually dependent on contextually dependent on the hiring economy. So if you're in a economy like we are right now where there is a lot more talent than jobs, then this

approach works great because you literally have like a stack of maybe hundreds of people and you want to like quickly weed out anybody that like you know doesn't you know doesn't want to meet up to like your very precise expectations. But if you're in the opposite, then there's way more of a selling of like, here's the way you would want to work here, here's all the great things.

And then the way most people work, just psychologically, is after they've signed, and maybe this is less true than it used to be, but after they've signed on and the bad things happen, they're more likely to kind of just keep sticking with it and deal with the bad things, then just go find a new job. No, that's not necessarily as true as it used to be. So I think historically though, like, all the like, you know, the rose, you know, kind of rose-colored glasses view of things, presenting that to the candidate did work because he landed people that maybe wouldn't have and they stayed at

least for some amount of time because like Yeah, sellers market for sure. That's yes. That's a really wise insight. Yeah. Yeah, but it was also I mean it did not help your team morale and stuff No, I mean, you know those were those places where it was like, yeah, you could get those people But then you could never let them interview a candidate because they'd immediately, like, their kind of bitterness or jadedness would go right off onto them. Yeah, and I think there's, to be clear, like, most people took it way too far. I think there's a, I think there's a version of it which is like, this is a misrepresentation of, like,

what you will do in this job that doesn't work for anybody ever. And then there's a version of it where we're not literally sending the person upfront every bad thing that could ever happen in the job and saying, are you okay with this? Because odds are it's not all gonna happen at once. I don't know, I don't think we've overcorrected too much. And obviously that post was a joke.

I don't think we've overcorrected too much. But there is a version where you can be so negative. You're like, I don't know if I want to work here. These people are weird. Like why are they super negative? Yeah. Yeah. Well, I think specifically what we used to see was a lot of like, I'm going to oversell what you're going to do in the role. Yeah. Particularly if it was like, cause you know, you get that it was the combination of people who were kind of like the, Hey, maybe this isn't going to pay as much, but here's all the cool stuff you're going to do,

right? And then you get in and it's like, yeah, can you manually process these 22 Excel files every day for us? Thanks. So that one, I think is one that really, I think got people really, did not like that. Well, that's worse. That's the worst thing to do. Cause the thing I've always sold people on more so than like company or money or anything is like, hey, if you do this role for a couple of years, you will be in a better spot career wise, regardless of where you work in the future. Like that's the best selling point.

Even if it's not the most fun job or the best company in the world, it's like, hey, do this for a couple of years, your career will progress. If you don't like it here, then leave. And you're still better off. Like that, that to me is the only thing that I've ever been able to consistently sell kind of regardless of company.

Yeah, I think when it comes down to it, building a team is just really hard. It is, you know, like keeping the team motivated, bringing new people on who are a great fit, especially when you're growing really quickly. It's just super hard, you know? It's super hard to do that. And then you get all of these things that are working against you,

including at least when you get to be larger companies, there's a lot of internal stuff that's working against you. Totally. You know, from, they can be HR departments, it can be working with, you know, internal recruiters. It can also just be that like, you can see a vision for where you want to go with this and you've got a boss or a boss's boss who's like, I don't really like that.

And I want to do something and I want to do this like stripped down version of whatever. And it's like, well, then why am I, you know, and so you're trying to balance all that stuff out of like, how do I move forward with what I feel and what I believe is the right way to do it? but also deal with the kind of the internal headwinds of people trying to Push back on that well, especially for technical leaders Where you get into the position you spend more and more of your time as you like, you know as your role increases if you're managing people or people or in director or executive levels, it's like, I just wanna build things.

Like I'm spending all my time clearing paths, dealing with people issues. And it's hard to find people that are willing to do all that and have a broad enough vision to realize they are still building things, but they're building things via proxy and still like get the satisfaction out of building via proxy versus tactile. Like I'm actually building with my own hands. And I think that is the really hard thing. And people tend to like, especially if you just love building, like shy away from like the actual work you need

to be doing and then just go build because that's fun and it's easier and it's comfortable. And you're not doing enough of the work around the, for your team. Well, and I think specific to that, I think the hard part is when, cause you can get the vision of like, I'm building, like I'm building a team and we're gonna build these things. And it's this like kind of a long-term thing

that we're doing. I think one of the hardest parts though, is most companies are really bad, matching up like, we need to discuss this first, we need to move and make a decision. And they tend to not be like, we always don't think, or we always overanalyze. It's more like we're really bad at matching the two up.

Like it's so frustrating on that, like kind of, I wanna build, I wanna do something when you're like, the answer is clear, gets so frustrating on that, like kind of, I want to build, I want to do something when you're like, the answer is clear, we know what it is, we just need to go like hire this person or do whatever. And then you get the like, well, I want to spend the next six months just talking about it. You're like, we don't need to talk about it.

And then it gets flipped on the other hand. And the next day it'll be like, by the way, we're making this massive change to the whole structure or environment or direction. And you're like, wait, oh, this is something that needs to be talked about. You can't do this. Yeah. It is funny how the same companies that spend six months talking about it will also do massive changes like without talking about it. Yeah. I've experienced that too. Yeah. And so I think that's one of the hard things about it is you feel like it kind of

hits you on both ends of where like you're not able to build in kind of like this the kind of like leadership way because everybody wants to relitigate everything and then suddenly they're just doing stuff and it's outside of your hands and you're like wait now I have to like fix this what just happened here yeah man I picked a great bonus round. Wow. I picked a great bonus round. If anyone listening has built a great data team, we would love to have you on the show. So let us know. And I think we're out of time. Yeah.

I'd also just say if you find anything particularly big on LinkedIn that you want us to talk about, just send it in, man. Do we have a way to do that, Eric? Maybe we can post something on the show notes like. Yeah, just email us. That'd be fun. Email Brooks at data stack show. Brooks at data stack show dot com.

Perfect. The data stack show or is it data stack show? I don't know. I don't even know our own website. That feels kind of important for this to work. I'm just going to throw that out. It is a small detail. OK, it is data stack show dot com. site. It's always great to have you and we'll catch you on the next one. Stay cynical. The Data Stack Show is brought to you by Rutter Stack. Learn more at ruddersack.com.