98: Day in the Life of a Data Analyst With Non-Technical Background w/ Paul Ahlstrom

[00:00:00] I was trying to help him work on his resume.

[00:00:01] And, but it's like, it's so hard when you're like starting completely fresh,

[00:00:05] like you have no data, anything except for portfolio projects.

[00:00:08] Yeah.

[00:00:09] She's got to find like that first person who can give you a leg in the door.

[00:00:13] Welcome to the Data Career Podcast, the podcast that helps aspiring data

[00:00:17] professionals land their next data job.

[00:00:20] Here's your host, Avery Smith.

[00:00:23] Data nerds.

[00:00:24] What's up you guys?

[00:00:25] Avery Smith back with another episode of the Data Career Podcast.

[00:00:28] And I'm really excited for this episode because I actually interviewed my

[00:00:31] childhood friend, Paul Ahlstrom, which was pretty fun because now he's in the data

[00:00:35] world and Paul has a history degree.

[00:00:38] He does not have a technical background and he was nice enough to sit down and

[00:00:41] kind of explain his journey into data analytics coming from a non-technical

[00:00:46] background, and he also talked about what the day in the life is for him.

[00:00:50] Like what tools is he using the most?

[00:00:52] What type of problems is he solving?

[00:00:54] So it's like a really applicable episode if you're like, well, what

[00:00:56] would life be as a data analyst?

[00:00:58] I think you'll really enjoy this episode.

[00:01:01] Just a few things before we get into this episode, I want to let you guys know

[00:01:05] that I have some great job tips that are coming out that are not necessarily just

[00:01:09] for data, but for all of job hunting.

[00:01:10] It's like basically a method to reframe the way you look at job hunting and

[00:01:16] hopefully make it easier, you know, less applications, more interviews type of a

[00:01:19] thing.

[00:01:20] And we have some, some free content that talks about that and I'll put the link

[00:01:24] towards that in the, in the show notes or the description down below.

[00:01:27] So check those out if you're in the job hunt right now, I think you'll find them

[00:01:30] useful.

[00:01:31] The other thing is we just released Avery GPT, which is basically like you take all

[00:01:35] of my knowledge out of my brain and throw it in a chat bot.

[00:01:39] And that is totally free.

[00:01:40] I don't even know why we're doing it, but it's totally free.

[00:01:42] Or basically you can get free coaching from me by asking whatever questions you'd

[00:01:47] like.

[00:01:47] So go check it out.

[00:01:49] Davecruiserjumpstart.com slash Avery GPT or we just have the link down below.

[00:01:52] Tell your friends about it.

[00:01:53] I think it's pretty cool and it's totally free.

[00:01:56] Most importantly, so you know why not take a look at it?

[00:02:00] I think those are the two main things right now.

[00:02:01] So we'll go ahead and get into the interview with Paul.

[00:02:09] Hey, I'm glad, I'm glad you're here, Paul.

[00:02:12] Thanks for coming on.

[00:02:13] Yeah.

[00:02:14] So Paul, for the listeners, we got to discuss how we know each other.

[00:02:18] I was trying to think when did we first meet?

[00:02:23] When did you start at Waterford?

[00:02:26] Third grade.

[00:02:26] Third grade.

[00:02:27] All right.

[00:02:27] So we've known each other since third grade.

[00:02:29] So since we were like, how old are you in third grade?

[00:02:31] Like 10 or something like that.

[00:02:34] Yeah.

[00:02:34] All right.

[00:02:34] And now we're obviously not 10.

[00:02:36] So we've known each other.

[00:02:37] We went to, I guess that would be elementary school, middle school, part of

[00:02:42] high school.

[00:02:43] I switched high schools after 10th grade.

[00:02:45] Did you graduate from Waterford?

[00:02:47] I moved to Mexico for most of high school.

[00:02:50] That's right.

[00:02:50] Came back to Sandy for the last year.

[00:02:54] So that's right.

[00:02:55] Yeah.

[00:02:55] So we didn't reconnect until later.

[00:02:58] Yeah.

[00:02:58] We actually are both members of the Church of Jesus Christ, Latter-day

[00:03:02] Saints, and we both served our two year missions in Sweden.

[00:03:05] So that's actually probably where we reconnected, which is crazy.

[00:03:09] Yeah.

[00:03:09] We were there at like the same time.

[00:03:10] It was kind of fun.

[00:03:11] I was there from October of 2013 to August of 2015.

[00:03:15] And you were there?

[00:03:16] Yep.

[00:03:16] I was six weeks after you.

[00:03:18] Oh wow.

[00:03:18] That's crazy.

[00:03:19] So that was the next transfer.

[00:03:20] Okay.

[00:03:20] And was Brock Johnson?

[00:03:22] Yep.

[00:03:22] Okay.

[00:03:23] This is super cool.

[00:03:24] Okay.

[00:03:24] This is super funny.

[00:03:25] Yeah, I was companions with Brock.

[00:03:26] Were you?

[00:03:27] Yeah.

[00:03:27] Okay.

[00:03:27] I was too.

[00:03:28] And now he works in data too.

[00:03:29] Well, that's what I was going to say is so, so anyway, so we're, we're

[00:03:33] missionaries and you know, we're hanging out basically, and you basically

[00:03:36] have what's called a companion, which is like the person that you are doing

[00:03:39] everything with, it's like a little built-in best friend or buddy.

[00:03:42] And so Brock Johnson was one of mine for six weeks, I think.

[00:03:46] And he was one of yours for about the same?

[00:03:48] For six weeks as well.

[00:03:48] Where was it?

[00:03:50] That was in Eskilstuna.

[00:03:51] Oh wait, I was, I was companions with Brock in Eskilstuna.

[00:03:54] No, no, no.

[00:03:55] I don't think you replaced me, but he was there for a long time.

[00:03:57] Okay.

[00:03:57] He was just there for a bit.

[00:03:58] Yeah.

[00:03:58] So that's super funny.

[00:04:00] So yeah, Brock and then Brock studied, I think it's like mechanical

[00:04:04] manufacturing engineering, and he worked at like a manufacturing company for a while.

[00:04:09] And then he tried doing some like window sales stuff and some other stuff.

[00:04:16] And then he moved to England and he reached out to me like, I don't know, a

[00:04:20] while ago, and he was like, I want to get into data.

[00:04:22] And so he actually joined my bootcamp.

[00:04:23] I don't know if you knew that.

[00:04:24] That's cool.

[00:04:24] I think he told me that.

[00:04:25] Yeah.

[00:04:26] So he was, he wasn't the member of the data analytic accelerator and was trying to

[00:04:30] land a job in England with no prior experience as a data analyst.

[00:04:36] So that one was tricky.

[00:04:37] Yeah.

[00:04:37] I was, I was helped.

[00:04:38] I was trying to help him work on his resume.

[00:04:40] And, but it's like, it's so hard when you're like starting completely fresh,

[00:04:43] like you have no data, anything except for portfolio projects.

[00:04:47] Yeah.

[00:04:47] Just got to find like that first person who can give you a leg in the door.

[00:04:51] Okay.

[00:04:52] And that's exactly great segue.

[00:04:54] That's exactly what I want to talk to you about because you're now a senior

[00:04:57] data analyst, Angel Studios, which for those who don't know, Angel Studios

[00:05:02] makes a lot of cool movies and television.

[00:05:05] Probably the most famous is The Chosen.

[00:05:08] What else do you guys make?

[00:05:09] Yeah.

[00:05:09] Angel Studios has changed its model a couple times throughout the years.

[00:05:12] It started as VidAngel, which was, you can filter movies and videos.

[00:05:17] Yep.

[00:05:18] And then it's recent, it's most current version of its business model is it does

[00:05:23] theatrical distribution for, well, distribution for movies and TV shows.

[00:05:30] So you can have an idea.

[00:05:32] They'll help you crowdfund your idea.

[00:05:33] They'll help you film, film it, distribute it, and do all the marketing and get it

[00:05:39] on, get in theaters, get it on platforms like Netflix.

[00:05:43] Oh, wow.

[00:05:44] So, so that's what, that's what you guys focus on now?

[00:05:46] Yeah.

[00:05:47] Oh, very cool.

[00:05:48] Okay.

[00:05:48] But if I look back on your resume, you know, eventually there was a point where

[00:05:52] you were financial analyst at Hamilton, you were data analytics engineer at

[00:05:55] Health Catalyst, but before that there's not a data analytics position.

[00:06:00] Yeah.

[00:06:00] So how did you get your foot in the door?

[00:06:03] Yeah.

[00:06:03] So about a year and a half ago I was a product manager and before that I

[00:06:08] managed a small video team and I had this feeling that I wanted to have a more

[00:06:14] technical skillset because I was, as a product manager, I was basically making

[00:06:18] to-do lists all day for developers.

[00:06:21] So making product decisions, which was fun and it was really interesting

[00:06:24] because that was the intersection of the user experience for a product and the.

[00:06:31] The development for, for that product to make the websites better.

[00:06:35] But I had this feeling like I really want to do something more technical and

[00:06:39] I want to do something that kind of feeds into my interests and I've always been

[00:06:41] interested in data stuff, but never had the vocabulary, the vocabulary.

[00:06:47] I never had the vocabulary or the skillset to act on that.

[00:06:52] So that's, that's where it started.

[00:06:54] Okay.

[00:06:54] And so you don't have the vocabulary, you don't have the skillset.

[00:06:57] You don't have the degree, right?

[00:06:59] Right.

[00:06:59] Because you studied, I looked at this here, history.

[00:07:02] Yeah.

[00:07:03] Okay.

[00:07:03] So you're studying history, which is like a non STEM, non-technical background.

[00:07:08] You're working as a product manager, which is like, I wouldn't say it's a

[00:07:11] technical role, but it's like you're getting closer than maybe history is a

[00:07:15] technical, technical thing.

[00:07:17] No, I didn't, I didn't get my history degree until after the health

[00:07:20] catalyst job, actually.

[00:07:21] Okay.

[00:07:22] So yeah, you don't even have a college degree at this point.

[00:07:24] So it's like, okay, why should anyone take it?

[00:07:27] Like, why should anyone hire Paul as a data analyst or how did you convince

[00:07:32] someone of that?

[00:07:33] Cause that's, that's something that's hard to do.

[00:07:34] Yeah, no, it is hard.

[00:07:36] So let's go back a couple hundred years.

[00:07:39] How did people used to get into jobs and careers?

[00:07:42] There used to be apprenticeships.

[00:07:44] You would say, I want to do this thing.

[00:07:46] You would get an apprenticeship in that field.

[00:07:49] You would learn the tools and the skills of the trade, and then you would become a

[00:07:52] master carpenter or whatever, after a couple of years of on the job experience.

[00:07:57] And that used to be the educational paradigm.

[00:08:00] Then we shifted to like universities where you come out with a degree, but with

[00:08:06] technology moving so fast, a degree doesn't tell you a whole lot about a person's

[00:08:11] skillset.

[00:08:12] So I can come in with a no degree or a history degree and people will still hire

[00:08:16] me for a data job.

[00:08:17] Why?

[00:08:18] Because they care about the specific individual unique skills that you've

[00:08:22] acquired and have gotten good at.

[00:08:26] It's the skills that matter.

[00:08:27] It's the skills that matter, but they have to be, people have to be able to trust

[00:08:30] that you, you have the skills and, and that's hard right now.

[00:08:33] There's a big burden on people that are hiring for analysts because they have to

[00:08:39] check every single person to see if they have the skills.

[00:08:41] And I think that's a whole, that's a gap in the, the job hiring marketplace right

[00:08:45] now.

[00:08:46] There's no, there's no good way for me to know.

[00:08:49] Oh, Avery is actually really good at SQL and he's really good at Python to like the

[00:08:54] 90th percentile or whatever.

[00:08:56] Yeah.

[00:08:56] I'd, I'd almost, I agree with you and I'd almost fight back with you on it a little

[00:09:01] bit because it's like, well, there's leak code.

[00:09:04] It's like, there's, there's like all these, like whenever I was interviewing for jobs

[00:09:08] for data scientist jobs, they'd send me the SQL assessments or these Python

[00:09:11] assessments.

[00:09:13] And so I'd almost argue with you.

[00:09:14] Well, no, there are things that test you, but I hate those.

[00:09:18] I hate those because you can't use Google now.

[00:09:21] Now I use chat GPT whenever I'm coding anything.

[00:09:24] You can't use those.

[00:09:25] And you have to do it in every single job interview.

[00:09:27] Right?

[00:09:27] Yeah.

[00:09:27] And it's like, and it's like you get like, like you have like a time limit and it's

[00:09:31] like super high pressure.

[00:09:33] Right.

[00:09:33] It's like, how do I do this thing in SQL that I like haven't done in like six

[00:09:37] months and remember without Googling it in like two minutes or less.

[00:09:41] Yeah.

[00:09:42] So I think, I think skills in general are going to become more anima atomized and

[00:09:47] then you're going to be able to get like a mini degree in a specific thing and,

[00:09:51] and, and prove demonstrate mastery in that one little specific thing.

[00:09:54] These, those micro skills, I think it's going to be a little bit like a tech tree

[00:09:57] like in video games where you like, you put like all your XP into, into like your,

[00:10:03] your spell casting or whatever.

[00:10:04] I don't play video games that often, but that's the analogy I think of it as where

[00:10:08] you, so like I have, if I've put all my XP into SQL for database for data analytics,

[00:10:15] right?

[00:10:15] Like, and I can, I'm like a level four SQL analyst.

[00:10:19] I think, I think that that's the direction.

[00:10:22] Like credentials and degrees are going to head.

[00:10:25] I, so yeah, I agree that there's, there's definitely something has to change in the

[00:10:29] future.

[00:10:30] Right now you have like these crappy.

[00:10:33] Skill assessments, which are hard to do and also just like not realistic because

[00:10:36] it's like in real life you get to use Google and real life you get to use chat

[00:10:40] GPT in real life.

[00:10:41] Like half the time, the, I don't know, I was always doing data science ones and so

[00:10:45] they were often Python and it was like calculates like the distance between this

[00:10:50] point and this point using sign.

[00:10:53] Like it was like always like, this is not realistic at all to what I'd be doing on

[00:10:57] the job.

[00:10:57] Yeah.

[00:10:58] So I think those assessments kind of stink.

[00:11:00] So regardless of where we're, where we're at with these not so good assessments,

[00:11:04] you still had, you mentioned that you not only have to have the skills, but you

[00:11:08] have to prove to someone else they have to be able to trust you that you have the

[00:11:11] skills.

[00:11:11] Right.

[00:11:11] Cause anyone can say that they know SQL, Excel, Tableau, Power BI on their resume.

[00:11:15] So how did you get someone to trust you?

[00:11:18] Yeah.

[00:11:18] So I got an internship or an apprenticeship at a company called Health Catalyst and

[00:11:26] the interview was basically, do you understand the concepts of data and SQL

[00:11:33] and like how to do joins, that kind of stuff.

[00:11:35] And once they could tell that I was kind of like smart enough to start the

[00:11:38] internship, they brought me on.

[00:11:40] It's their pipeline for four new analysts or analytics engineers or analysts.

[00:11:47] So here's a company who's going back to the apprentice model in some, in some

[00:11:53] regards, and they're saying, well, you're smart enough.

[00:11:55] We will teach you everything you need to know to do this job.

[00:11:58] Yeah.

[00:11:59] And I, first off, I think that's such a cool way to like actually break into the

[00:12:03] fields.

[00:12:04] I don't think there's many companies doing it.

[00:12:06] I think Health Catalyst is one.

[00:12:07] Another one I really like, we're going to talk about more about education in this

[00:12:10] podcast, but another one I really like is called the Data School.

[00:12:14] Have you ever heard of it?

[00:12:14] No.

[00:12:15] It's in, it's only in, they have it like in New York, London, Germany, like, but

[00:12:19] the only one in the U S is in New York city.

[00:12:21] It's really cool.

[00:12:22] We had one of the co-founders on an episode, like 73, I think, I think right

[00:12:30] around there.

[00:12:30] So definitely, definitely worth looking.

[00:12:33] We'll have a link in the description down below.

[00:12:34] So hear that.

[00:12:35] But the Data School is not actually a school.

[00:12:37] It's an apprenticeship program where they take people who want to transfer into

[00:12:41] data.

[00:12:42] The interview process is pretty, pretty lengthy and they go through actually

[00:12:45] like you have to make a custom project for them and present it.

[00:12:49] And they hire about like eight people, a quarter or something like that.

[00:12:53] And it ends up being like $65,000 a year.

[00:12:57] And you are now a data consultant and they're a consulting company and they're

[00:13:01] like working with a bunch of clients and customers and you get to be, you know,

[00:13:04] you get, you basically get, I think it's like an 18 month program.

[00:13:08] The first like six months you're learning the next six months you're like

[00:13:11] practitioning the last six months.

[00:13:12] You're kind of like leading some of the newer people.

[00:13:14] And it's like, that's how that's an apprenticeship program.

[00:13:17] I don't think there's many of those.

[00:13:19] No, but I think they're cool.

[00:13:21] I think that's super cool.

[00:13:22] I think what you're doing is really awesome too.

[00:13:23] Like just it's, it's the new way to get educated.

[00:13:27] It's like you, you, you start like a specific program about something that is

[00:13:32] going to prove that it has value and you're going to end up with at least

[00:13:37] projects and demonstrable experience.

[00:13:40] But yeah, I think that's the way education definitely needs to head.

[00:13:43] Which is what I was going to say.

[00:13:44] It's like, if you can't find one of these apprenticeships, doing projects is

[00:13:48] probably the next closest thing where it's like you're giving your, you're

[00:13:51] hiring yourself to solve a problem that you want to solve, like that you've an

[00:13:57] interesting thing or something.

[00:13:58] You can always pretend that you work for a certain company or something like that.

[00:14:02] But I want to ask you, like, how did you with health catalyst?

[00:14:05] Cause that's like a great place to find this apprenticeship program.

[00:14:09] How'd you even know it existed?

[00:14:10] Cause, cause I've, I actually applied to health catalyst when I was in college

[00:14:15] and I got rejected health catalyst.

[00:14:17] If you're listening, I'm not, I'm not missed out.

[00:14:20] He didn't miss out, but I probably would have quit and left you anyways for bigger

[00:14:24] and better things, I guess.

[00:14:25] But my point is like, I didn't even know that program existed.

[00:14:28] Maybe, I don't know.

[00:14:29] I don't know.

[00:14:29] I don't even know that.

[00:14:30] How do you know that program existed?

[00:14:31] Yeah, it was someone or it was, yeah, it was an introduction.

[00:14:33] So I went to a friend's house and his uncle founded health catalysts and he

[00:14:38] was like, oh, we have a really cool program.

[00:14:42] And I was like, that sounds great.

[00:14:44] I don't even know what it is.

[00:14:45] So you, you basically, your networking wasn't even really, your networking led to

[00:14:52] this, you know, interview process where you still had to ace the interview.

[00:14:55] You still had to do well, but you would never would have even had the interview

[00:14:59] had you not randomly known this person.

[00:15:01] And the cool thing is this, it's not like you were trying to network really.

[00:15:05] Right.

[00:15:05] Like it's just like a friend and you were just like around the

[00:15:07] friend's family or something like that.

[00:15:09] Yeah.

[00:15:10] The important starting point for me was knowing what I wanted to look for, even

[00:15:14] if I didn't have like a total, like a name for it.

[00:15:17] So I knew I, I really like being able to help people solve problems.

[00:15:21] I love looking.

[00:15:23] I love being able to solve process problems and look for improvements.

[00:15:27] And I like using data.

[00:15:29] I don't have the technical skills and I want more technical

[00:15:31] skills to do that sort of thing.

[00:15:32] So I started from understanding my interests and what I was good at,

[00:15:37] and then felt it out from there.

[00:15:40] Yeah.

[00:15:40] I that's so key because when I was just trying to be a lab technician back in the

[00:15:46] day, I was, I was just in college.

[00:15:48] I was excited to, you know, be doing, having a big boy job and stuff.

[00:15:52] And I was cold messaging everyone trying to get a lab technician job and no one

[00:15:57] would take a chance on me because I didn't have any experience.

[00:16:00] I ended up getting an opportunity that paid absolutely terribly.

[00:16:04] And I worked there for three months and I got some experience, but it wasn't until

[00:16:10] I went to church and this isn't like, I went to my, basically my parents' church.

[00:16:14] I went back home for Christmas, went to my parents' church and I ran into one of my

[00:16:19] old church leaders and he was like, Oh, what are you doing nowadays?

[00:16:22] And I was like, Oh, I'm a lab technician.

[00:16:24] And he's like, Oh, well, did you know that I work at a lab?

[00:16:28] No.

[00:16:29] What lab do you work at?

[00:16:30] Oh, it's actually, you know, really close to where you live.

[00:16:33] Oh, oh my gosh.

[00:16:34] Well, that's funny.

[00:16:36] Would you want to interview there?

[00:16:36] Yeah.

[00:16:37] Okay, sure.

[00:16:38] And then it was actually at that job where I ended up learning to become a data

[00:16:41] analyst and then data scientist and like, I owe everything to that random

[00:16:46] conversation at church.

[00:16:47] But I think sometimes people get really nervous about networking because it's

[00:16:51] like, Oh, I have to like go to networking events or I have to post a lot on

[00:16:55] LinkedIn, which I think are both two great things to do.

[00:16:58] But in your case, it was like, no, you hang out with a friend and talk about

[00:17:01] your dreams and your job and your career and stuff like that.

[00:17:03] Yeah.

[00:17:04] I just told him like, I'm looking for a job change.

[00:17:06] He's just like, he was like, what do you want to do?

[00:17:07] I was like, I want to like learn data stuff and I want to do process

[00:17:11] improvements.

[00:17:11] He's like, well, I had the internship for you.

[00:17:14] It's not going to happen like that for everyone.

[00:17:16] It's not.

[00:17:17] But my, I have a fam, I have a close family member who owns a recruiting

[00:17:21] company and their stats are if you, if you go the cold application route for

[00:17:26] most things, even if you have experience in that industry, it's going to take

[00:17:29] you 200 job applications or so before you get the job offer.

[00:17:34] I have a friend who has a couple of years of experience who just did 600 and

[00:17:38] got two job offers.

[00:17:39] So 300 to one.

[00:17:41] Yeah.

[00:17:42] And also in data.

[00:17:44] Yeah.

[00:17:44] And I think the statistics on networking, it's, it's a lot lower.

[00:17:51] It's like 30.

[00:17:52] Yeah.

[00:17:53] Or even less because you, you can bypass their applicant tracking systems and you

[00:17:59] can, you come in as a known entity.

[00:18:01] This is a, this is a good person who I trust.

[00:18:04] And then all you have to do from there is demonstrate your skillset.

[00:18:06] Yeah.

[00:18:07] Which, which is like, it seems so simple, but it's so impactful.

[00:18:12] I love that data set.

[00:18:13] I have a data set that is not nearly as probably as legitimate.

[00:18:17] It's a LinkedIn survey.

[00:18:19] But one time my friend did a LinkedIn survey and they asked, how did you, or

[00:18:22] how do you approach the job hunt?

[00:18:25] Yeah.

[00:18:25] And I think around 66% of people said they find jobs online and they apply.

[00:18:31] And then the next day he asked, how did you get your last job?

[00:18:34] And only 33% had done it through cold applying and the remaining 66 were

[00:18:40] through getting recruited and referred.

[00:18:41] Yeah.

[00:18:41] And so I kind of call it the job hunt Pareto principle.

[00:18:44] Basically you're going to spend 80% of your time applying for jobs, but only

[00:18:49] only those only lead to 20% of actual hiring when you really should be spending

[00:18:53] 80% of your time networking.

[00:18:55] Cause that's going to lead to 80% of the actual job hires.

[00:18:58] So I just think that's really key that like you had this goal to like get there

[00:19:03] and you're like, how do I get there?

[00:19:04] And the answer was you got a network basically.

[00:19:07] Yeah.

[00:19:07] And then show off your skills because you couldn't just be like a nobody.

[00:19:10] Like a bum, like you had to know at least some SQL and stuff like that.

[00:19:14] Yeah.

[00:19:15] Okay.

[00:19:15] Very cool.

[00:19:16] And, and that was an analytics engineer job.

[00:19:18] Is that right?

[00:19:19] Yeah.

[00:19:19] Analytics engineer.

[00:19:20] So they, they taught me everything I needed to know in terms of both SQL

[00:19:24] and their internal tools.

[00:19:26] Okay.

[00:19:26] And can you speak a little bit to what an analytics engineer does versus like a,

[00:19:32] like a data analyst type of a thing?

[00:19:34] Sure.

[00:19:35] Yeah.

[00:19:35] This industry is, is slowly developing its naming convention.

[00:19:39] So like what is a data engineer versus data analysts versus an analytics engineer?

[00:19:43] Analytics engineers handled the transformation step mainly.

[00:19:48] So they will, they have, they start with the source data and then use tools like

[00:19:54] dbt or in health catalyst case, they have an internal tool for that to do all the

[00:19:59] transformation steps to create data warehouses, data marts for businesses,

[00:20:04] for specific use cases, the analytics engineers at health catalysts are also

[00:20:08] analysts.

[00:20:08] So they're, they're building out power BI dashboards and doing a bunch of SQL

[00:20:12] transformations to create the useful tables.

[00:20:14] So it's, it's somewhere between a data engineer and a, an analyst.

[00:20:19] Yeah.

[00:20:19] It's almost like a bridges the gap between a data engineer and the data analyst.

[00:20:23] And sometimes they do a data engineer's job.

[00:20:27] And sometimes like you said, they do a data analyst.

[00:20:28] Yeah.

[00:20:29] All the naming conventions are pretty fuzzy.

[00:20:31] And I think the more important thing is like, what are you actually doing with your,

[00:20:35] with your time?

[00:20:35] Yeah.

[00:20:36] Yeah.

[00:20:36] A hundred percent.

[00:20:37] So you did that and then you were a financial analyst and now you're a senior

[00:20:40] data analyst.

[00:20:42] Do those roles, do they have like a lot of overlap?

[00:20:44] Like what's, what's something that you've done at all three of those roles and

[00:20:46] what's something maybe you only did at one?

[00:20:48] Done a lot of SQL at all three of those roles.

[00:20:51] And I've done a lot of data visualization on power BI or the first two and a

[00:20:57] metabase at the third.

[00:20:59] So digging through the databases, making sure the data is clean, that you, you

[00:21:05] know, check all your assumptions, do all your filtering, pass it into a BI tool.

[00:21:10] And that's, that's what I've done in all three jobs.

[00:21:13] Okay.

[00:21:14] Let's dive in a little bit deeper to what you're currently doing, but, but you work

[00:21:18] at Angel studios, senior data analyst.

[00:21:20] This, this company is basically in the, you explained it kind of like the, we'll call

[00:21:25] it the show television movie, get stuff into production business and marketing

[00:21:32] distribution, marketing distribution.

[00:21:34] Do they have their own platform?

[00:21:36] Yeah.

[00:21:36] Okay.

[00:21:36] You can go to angel.com and watch some, some of our shows.

[00:21:39] So it's kind of like Netflix or Hulu, but a little bit smaller, obviously.

[00:21:43] So what, when you've been there, what are some, some of the

[00:21:45] problems you've been tackling?

[00:21:48] Most of my time is spent doing ad hoc requests.

[00:21:51] Really?

[00:21:51] So people need an email list for something very specific.

[00:21:55] So like users who watch this TV show did this behavior on our app and, and have

[00:22:02] watched this other thing more than once.

[00:22:04] Like get a lot of user user behavior kind of data requests.

[00:22:09] Super interesting.

[00:22:10] So someone will be like, for whatever purpose they want to segment.

[00:22:14] Part of their email list.

[00:22:16] Uh, and I was thinking, well, why can't you just do that in like

[00:22:20] whatever email tool you're using?

[00:22:21] Like I do that with the email tools.

[00:22:23] I said not like, okay, this person clicked on this, this person clicked on that.

[00:22:26] But my email software doesn't have any data about like what they

[00:22:30] watched, right?

[00:22:31] And your guys's platform.

[00:22:32] So that's why you have to go to like a lower data level to solve that problem.

[00:22:36] Right.

[00:22:36] Unless you have like a complex, this business has been through a lot of

[00:22:40] changes, so they don't have like established like processes for every

[00:22:43] single thing.

[00:22:44] So like let's tag this user as a, as a watch, like someone who's watched this

[00:22:48] show.

[00:22:49] So it instead they have a flexible data warehouse and then they have an analyst

[00:22:53] like me who can pull out answers to those types of questions.

[00:22:57] And then eventually once, once those processes are more established, they

[00:23:01] won't need me anymore.

[00:23:02] And I can, I can, you know, I'll be focused on other parts of the business.

[00:23:05] So just fire you and use chat.

[00:23:06] No, they'll be using me to answer more, more pertinent questions.

[00:23:11] Okay.

[00:23:11] Okay.

[00:23:12] So you're doing like a lot of ad hoc stuff and is that kind of like in sequel

[00:23:15] ish types type languages?

[00:23:16] Yep.

[00:23:17] I spend 90% of my day in SQL.

[00:23:18] Wow.

[00:23:19] Really?

[00:23:20] Yep.

[00:23:20] Just doing a bunch of queries.

[00:23:21] Okay.

[00:23:21] So, so like that's, that's like a business use case of they need an email list.

[00:23:25] What's like another business use case that you could share with us?

[00:23:29] Doing reporting for, for making business decisions.

[00:23:32] So right now we're working on distributing a new, a new film and they need to know

[00:23:36] where are we going to put billboards?

[00:23:38] Where are we going to pay for traditional media that's not connected to social media.

[00:23:44] And so I'm working on some more advanced analytics to pull out information about

[00:23:48] showtimes information about people who have purchased tickets in those areas and

[00:23:52] combine that into a useful report for making multimillion dollar decisions.

[00:23:58] Wow.

[00:23:59] Which is a lot of pressure.

[00:23:59] That is a lot of pressure.

[00:24:01] That's how do you feel?

[00:24:02] It's like, I gotta make sure I use the right wear filters.

[00:24:05] Yeah.

[00:24:05] That's a, if you make a, yeah, you have something wrong in that query.

[00:24:09] That could be.

[00:24:10] Yeah.

[00:24:11] It's like, yeah, we should have billboards in South Dakota.

[00:24:14] Yep.

[00:24:15] Yeah.

[00:24:15] And then it just turns out to be like some like random data problem that you didn't

[00:24:18] check.

[00:24:19] So there's a lot of things you need.

[00:24:21] You need to be careful.

[00:24:22] Okay.

[00:24:23] As a, as a habit.

[00:24:24] Just what you just mentioned where to put billboards, traditional media.

[00:24:28] I'm assuming that's like, like things like the today show or, or, you know, like

[00:24:33] guests on like Jimmy Fallon or something like that.

[00:24:35] Right.

[00:24:36] That's that to make those types of decisions to make data-driven decisions

[00:24:40] with, with that, you'd have to have data surrounding that.

[00:24:43] So are you guys using, I'm assuming you guys don't have all that in-house.

[00:24:46] You, you have to be using some external data as well.

[00:24:49] Yeah.

[00:24:50] A couple of external data sources, but it's mostly like, so like with, with, with,

[00:24:55] with digital advertising, you can say, Hey, Facebook, here are all the users that

[00:25:00] have clicked on this button.

[00:25:01] And then Facebook will go and figure out like using their, their in-house AI stuff

[00:25:06] exactly who else they should target.

[00:25:08] It looks like those other people when you're making decisions outside of those ad

[00:25:13] platforms, you have to kind of DIY it.

[00:25:15] So you're pulling data from various sources.

[00:25:17] You're, you're combining it in new and unique ways.

[00:25:21] And you're, you're starting with your own intuition about how you're going to solve

[00:25:24] that problem.

[00:25:25] Okay.

[00:25:26] Yeah, that, that makes a lot of sense.

[00:25:27] And yeah, we really don't realize how much data is necessarily being stored on us.

[00:25:34] So like for, for example, if you're scrolling through Netflix, that scroll is

[00:25:39] probably being stored one way or another.

[00:25:41] Yeah.

[00:25:41] You pause on a thumbnail that's being, that's being stored and that data is

[00:25:45] available to make data-driven decisions.

[00:25:47] So you open up an app, the time that you open up the app that's stored.

[00:25:51] It's all, so yeah, every, every event you, every user interaction event is, is being

[00:25:56] sent to a database where an analyst like me can poke around and try and figure out

[00:26:02] what you did and why and when.

[00:26:03] Where, where Paul's stalking you basically.

[00:26:05] Yeah, basically.

[00:26:06] Which, which like people, I feel like most people are like, oh, that's scary.

[00:26:11] Big brother.

[00:26:12] I don't like that.

[00:26:13] But as a data analyst, I've always loved it.

[00:26:15] I think that's so fascinating.

[00:26:17] Yeah, it is cool.

[00:26:18] Yeah.

[00:26:18] Okay.

[00:26:19] So that's kind of what you're working on 90% of your day.

[00:26:21] In SQL.

[00:26:22] Yeah, basically.

[00:26:23] Okay.

[00:26:24] So if you're not in, let's say, let's say you're not crunching numbers at your job.

[00:26:29] What are you doing?

[00:26:31] Yeah.

[00:26:31] So I'm, I'm in meetings with stakeholders around the business talking about how we're

[00:26:36] going to structure our data warehouse, how we're going to use our different tools to

[00:26:41] collect and store and transform data.

[00:26:44] So I'm, I'm having a, I'm playing a pretty big part in helping make decisions about,

[00:26:49] or like what, where are we going to store all this stuff and how, what format are we

[00:26:53] going to store it in?

[00:26:54] And, and then what tools are we going to use to transform it?

[00:26:57] And how is it going to be served in the end to end users?

[00:27:01] That makes sense.

[00:27:02] So you're, you're talking to your team about like, how are we even going to do this

[00:27:06] whole data thing?

[00:27:07] Cause you guys, like you said, are kind of a young company and you've had a lot of,

[00:27:11] you've switched the company has switched what they've done a lot in the past and

[00:27:15] ownerships and stuff like that.

[00:27:17] Yeah.

[00:27:17] So talking about data governance initiatives.

[00:27:19] So how, like where, where can I find these tables and who owns these tables?

[00:27:23] What's the best way for us to work together as a team from both the data producer and

[00:27:27] consumer side to make sure we're all rowing together as a company and making data

[00:27:33] informed decisions.

[00:27:34] We're talking about like the analytics engineering side of it.

[00:27:38] So like how, how can we transform this raw data into something super useful for

[00:27:43] downstream, downstream consumption?

[00:27:45] I'm talking about data modeling and how we're going to put everything in the

[00:27:49] warehouse so that it can be the most useful.

[00:27:53] Wow.

[00:27:53] That kind of stuff.

[00:27:54] Okay.

[00:27:55] And I want to go back one step and just hear what you have to say about your

[00:28:01] reporting.

[00:28:02] So you get these ad hoc requests, you're crunching the numbers in SQL.

[00:28:06] How, how do you tell them the results in SQL?

[00:28:10] Like are you, are you just, is this just all via email?

[00:28:13] Are you, are you meeting?

[00:28:14] Are you making a PowerPoint?

[00:28:16] Are you just screenshotting your SQL output and texting it to them?

[00:28:21] Sometimes I will send screenshots.

[00:28:24] I'm a firm believer that you never optimize early.

[00:28:27] So if someone just needs a list, I send them a CSV file and I tell them, I send

[00:28:32] them like the query and explain what it does.

[00:28:34] Okay.

[00:28:34] So you're exporting to a CSV from SQL or from whatever platform.

[00:28:38] Yeah.

[00:28:39] If it's an engineer that needs a query or something, then I'll just send them the

[00:28:42] query and they can run it themselves.

[00:28:44] If it's someone who needs a dashboard, I'll, I'll build the dashboards and the

[00:28:47] link to it and then tell them how I, how I made it.

[00:28:50] And if they need more explanation, we can have a conversation, a phone call or

[00:28:54] something.

[00:28:55] Okay.

[00:28:55] So you're either, you're either sending like messages on like an internal chatting

[00:29:00] platform or your, I'm assuming that's what you might do with the queries even.

[00:29:04] Maybe, maybe that's a lot of back and forth on Slack.

[00:29:06] Okay.

[00:29:07] There you go.

[00:29:07] Okay.

[00:29:08] Or you're just sending them a link to a dashboard.

[00:29:11] Yep.

[00:29:12] Okay.

[00:29:13] I like that a lot.

[00:29:14] Yeah.

[00:29:14] But the, the most important part of that whole process is the requirements

[00:29:18] gathering step.

[00:29:19] So you don't, you don't necessarily need to have a lot of meetings after you

[00:29:23] report, but you do need to have a lot at the beginning to make sure you understand

[00:29:27] the requirements because if you, if you don't ask the right questions about what

[00:29:31] they're really asking you for, then you're going to spend the next eight hours

[00:29:36] building the wrong thing because you didn't check your assumptions on, on what

[00:29:39] they, what they really needed.

[00:29:41] Yeah.

[00:29:41] And I guess if you're not clear at the beginning too, you could also be making,

[00:29:45] giving them the complete wrong email list.

[00:29:48] Right.

[00:29:49] It could be like, Hey, I'd need, this is from, this is too dumb of an example,

[00:29:53] but if they're like, Hey, I need all the people who watched Barbie, but you send

[00:29:55] them to people who watched Oppenheimer, that's probably not the right list.

[00:29:58] Yeah.

[00:29:59] Or like, like, like for example, like if, if who they consider to be someone

[00:30:04] that, that watched something, cause you might just take, okay, I'll take anyone

[00:30:08] that watched this thing, no problem.

[00:30:09] Send it to him.

[00:30:10] And then you find out, Oh no.

[00:30:11] Like we consider a viewer, someone that watched something for more than 30

[00:30:14] seconds.

[00:30:15] So there's, there's business definitions that you can't just assume implicitly

[00:30:20] or else you're going to get it wrong and you might, you know, make a mistake.

[00:30:23] Yeah.

[00:30:24] That's a, that's a tough meeting to have.

[00:30:26] Yeah.

[00:30:26] And so it's better to be careful.

[00:30:28] Ask a couple of follow-up questions like what are you using?

[00:30:31] And the best question you can ask is what do you need it for?

[00:30:34] And, or can you give me the background and the context on this, on this thing?

[00:30:37] What are you trying to do?

[00:30:39] Because often they, they might be asking you for one thing.

[00:30:42] But they need another, and you're the only one that knows that because you're

[00:30:46] the one digging around in the data all day.

[00:30:48] So if they need to know, you know, viewership statistics and you have

[00:30:52] another data source that would be better for that, it's your job to be like, okay,

[00:30:56] so you're at, are you asking it for it for this reason or for this reason?

[00:31:00] Cause it's, if it's for this reason, then let's look at this data.

[00:31:04] So that's one of your responsibilities in your data analyst role is to actually

[00:31:09] figure out what they actually want.

[00:31:11] Yeah.

[00:31:11] It's your job to properly, you, you are a free agent.

[00:31:15] You're not just like a SQL monkey.

[00:31:17] You are, your job is to interpret the request and then decide on the best

[00:31:22] path forward to get to the right answer.

[00:31:25] I wonder if that's something you've kind of developed as you've gone though,

[00:31:28] because it seems like when you're first getting started, that might be a big step.

[00:31:33] Sure, sure.

[00:31:34] Well, yeah, you gotta, when you're first getting started, your first

[00:31:38] question is where do I find that?

[00:31:41] What table does that live in?

[00:31:42] Yeah.

[00:31:42] And then your next question is like, which columns should I use exactly?

[00:31:46] Yeah.

[00:31:46] And then you're, and then, but the, the, the really good thing I learned from

[00:31:51] health catalyst is that great analytics are developed iteratively.

[00:31:55] Hmm.

[00:31:56] So if you just say, okay, I did a query, here's the report.

[00:32:03] You don't get to like the real answer that they were looking for.

[00:32:08] So I spent all my time at health catalyst working on one dashboard for a team.

[00:32:12] And it was like this, like they said, we need this put in the power BI dashboard.

[00:32:18] I would go and figure out how to do that.

[00:32:20] Deliver the metric.

[00:32:21] And they said, okay, it looks good, but let's make it like this instead.

[00:32:24] I would go through another round of iteration after like 150 of those little

[00:32:29] cycles, you, you end up with a really amazing dashboard.

[00:32:33] They were like, wow, it looks like like a professional with experience did this.

[00:32:37] And this is one of the best examples of analytics I've seen at health

[00:32:41] catalyst, which is high praise for a analytics company.

[00:32:44] They do analytics consulting.

[00:32:45] So they have like a hundred, 200 and it wasn't because I'm like smart or anything.

[00:32:50] It was because we did a ton of iterative cycles for one specific set of questions.

[00:32:55] They wanted to know how can we allocate engineers to our customers and who's

[00:33:01] allocated fully, who's under allocated, who's over allocated, who's working too

[00:33:05] many hours, who's not working enough hours.

[00:33:07] And as we iteratively through a lot, a lot of messages over and over again,

[00:33:13] refine refining things and combining things together, you started to get

[00:33:17] more and more useful insights.

[00:33:20] I love that.

[00:33:20] And for our wonderful video audience that is watching, we're in a new studio, by

[00:33:26] the way, if you can tell we're in person as well, we're saying that like 40 minutes

[00:33:29] in, but hopefully you guys can see it.

[00:33:31] So if you're watching on YouTube or on Spotify, what Paul just did was like,

[00:33:36] move his hands like this, like a hundred times in a row.

[00:33:38] And I love that because I think I love what you said that good data analysts are

[00:33:44] iterative and good data analytics is iterative.

[00:33:47] You keep, you do something, you talk to your shareholder, it goes back to you.

[00:33:52] You go back to your shareholder and you just kind of keep doing that until, you

[00:33:56] know, you've done it so much that it's like, Oh wow, then we actually made

[00:33:58] something pretty great.

[00:34:00] A hundred percent.

[00:34:00] That's how it, like a junior person like me, like my first data job made actually

[00:34:05] made something that was like super useful for a senior VP.

[00:34:08] And they were like, they were crazy excited about it.

[00:34:10] And it wasn't because I was smart or because I was like good at Power BI or

[00:34:13] SQL, cause I sucked at both at the beginning.

[00:34:15] But we did a ton of iteration on it and they were willing to spend the time with

[00:34:19] me to make those dashboards great and get to the answers and check the data with

[00:34:23] me.

[00:34:24] And being able to do that if you had the time and bandwidth is, is very powerful.

[00:34:29] Which is, which really goes along of today's theme of the podcast, which is,

[00:34:33] guess what?

[00:34:35] It's not really your skills that gets you hired.

[00:34:37] Usually it's your network that gets you hired.

[00:34:39] Yeah.

[00:34:40] And guess what?

[00:34:40] It's not your skills that impress the senior vice president.

[00:34:44] It's really your communication and like your, your ability to present and

[00:34:49] communicate with them effectively.

[00:34:51] Yeah.

[00:34:52] I, I think I, I definitely almost over communicated in that role because I was,

[00:34:57] it was, it was just me and him.

[00:34:58] I didn't, I wasn't working on a team or anything.

[00:35:00] So I was just like all day, like, Hey, this is right.

[00:35:02] Is this right?

[00:35:03] And he was patient enough with me to be like, looks good.

[00:35:05] Let's fix this.

[00:35:06] I think another thing that's super important is to have a bias for, for

[00:35:09] action instead of sitting on a problem and not knowing the right answer.

[00:35:15] So if you can deliver a kind of bad version of something quickly, and then

[00:35:19] iterate a couple of times, that's better than taking like six weeks to come to a,

[00:35:23] an okay answer.

[00:35:25] A hundred percent because most of the time, most of the time I've seen this on

[00:35:30] both sides of the aisle now where I, I used to be, you know, the person that was

[00:35:34] reporting to stakeholder.

[00:35:37] Right.

[00:35:38] But now as I'm transitioned and I try to do more CEO things, I've had people, you

[00:35:43] know, that report to me and I've seen it from both sides and I've realized that

[00:35:47] neither side actually knows what they want or what they're doing until like three

[00:35:51] iterations in.

[00:35:52] Right.

[00:35:53] Cause it's like, it's actually one of the things you mentioned earlier was you ask

[00:35:57] the key stakeholder, well, what do you, what do you want this for?

[00:36:01] What are you trying to do?

[00:36:02] And asking that question is so important because that key stakeholder isn't even a

[00:36:06] hundred percent sure.

[00:36:07] They're like, Oh, you're right.

[00:36:09] What do I want this for?

[00:36:09] Oh, it's for this.

[00:36:10] Well, is it actually for this or is it for that?

[00:36:12] And you're actually helping that person, you know, create their, their vision of

[00:36:16] the tool.

[00:36:17] And so sometimes this, the, these busy managers, they can't actually envision

[00:36:21] what they want until you've given them something and they're like, Oh, this is

[00:36:24] great.

[00:36:24] It's exactly what I want.

[00:36:25] Or no, you're totally off.

[00:36:27] It's not what I want at all.

[00:36:28] Yeah.

[00:36:29] And if, if you, if you just go off of what they ask for, you're going to get fired

[00:36:33] soon because I like that because.

[00:36:36] You, you need to, you need to, you can't play the game of guessing what's in

[00:36:40] their head.

[00:36:40] You just have to like, like drill down and figure out what's actually going to make

[00:36:44] you useful and make your data useful.

[00:36:47] Because if, if they ask for this thing and you give it to them and then it's like

[00:36:50] not that helpful.

[00:36:52] What if you instead ask them a couple of times, well, why do you need that?

[00:36:55] Oh, I need it for this.

[00:36:56] Okay.

[00:36:56] Well, what if we did this instead or combine this data?

[00:36:58] Oh yeah.

[00:36:59] I didn't even realize that was possible.

[00:37:01] That would be so helpful.

[00:37:03] If you can, if you can go through those iteration cycles and then always be like

[00:37:07] questioning, like, okay, well, why do you really need that?

[00:37:10] And wouldn't this be a better way to do it?

[00:37:12] And you'll run into a deeper understanding of the business logic behind like what

[00:37:17] you're trying to accomplish and get to something more useful, faster.

[00:37:21] A really interesting, cool concept.

[00:37:24] We just interviewed Nick Debra, who's data visualization expert.

[00:37:28] And one of the things that he coined is spray and pray data visualization.

[00:37:32] And basically when you're working with a business share or I could never say that

[00:37:36] right.

[00:37:37] Shareholders, stakeholders, stakeholders, probably shareholders to the early on a

[00:37:41] little bit of shares, stakeholders.

[00:37:43] And they ask you to create a visualization and you're not, you're like, what, what am

[00:37:46] I supposed to do?

[00:37:47] And they're like, I'm busy.

[00:37:48] Go do it.

[00:37:48] You know, his approach is to like, all right, here's four visualizations.

[00:37:53] Which one you like choose, choose one of these four, which one resonates with you.

[00:37:57] And it's almost seen, seems like you're suggesting something similar where it's

[00:38:00] like when they ask you to do something, you're like, here are four interesting

[00:38:04] questions that we can kind of take this analysis.

[00:38:07] Which question, you know, is the most useful to you or like which one you're

[00:38:11] drawn to?

[00:38:12] Spray and pray.

[00:38:13] It sounds like evolution, like evolution creates like 10 different versions of

[00:38:17] something and then one or two survive.

[00:38:18] Exactly.

[00:38:19] Cause most dashboards don't survive.

[00:38:20] No.

[00:38:22] Most, most dashboards I've created are like, or oftentimes you'll create something

[00:38:26] and then it'll get used once or not at all.

[00:38:28] And you know, cause if, and that's a signal to you as an analyst that you didn't

[00:38:32] really create something that they needed.

[00:38:34] Hmm.

[00:38:35] Interesting.

[00:38:36] I like that.

[00:38:37] I could see that.

[00:38:38] Could it also be that you didn't, they don't know how to use it or like they

[00:38:41] don't know where it is or it's not accessible.

[00:38:43] Also could be the case.

[00:38:44] I saw an interesting thread the other day on like on a CEO wanted a dashboard like

[00:38:50] sent to his email and this person was like criticizing the CEO, like, Oh, like you're

[00:38:55] wasting engineering time, but that comes from a misunderstanding of the context of

[00:39:00] your user, which is the CEO in this case where he, he doesn't have time to log into

[00:39:04] some platform and like, and use his credentials to see the dashboard.

[00:39:09] He just needs the numbers or the answer like where he is right now.

[00:39:13] And so understanding the context of your user is also an important part of being

[00:39:18] a super useful.

[00:39:18] Cause if you make a dashboard and they don't use it, then you might as well

[00:39:21] have not made the dashboard.

[00:39:23] I want to talk about something that you just mentioned, which was, yeah, this is,

[00:39:26] this is how you could get fired quickly.

[00:39:29] I think one thing that you are qualified to speak on is how do you get, how do you

[00:39:34] get promoted quickly?

[00:39:35] And I'll make, that might be a hard question, but one of the things that I think that

[00:39:40] one of the reasons why I think you'd get promoted or you will, or you have gotten

[00:39:43] promoted is what you just mentioned.

[00:39:45] The action, the bias to action.

[00:39:47] Yeah.

[00:39:47] To be promoted in a company, your interests have to be aligned with what

[00:39:52] the business is trying to do.

[00:39:55] So the business is trying to make money.

[00:39:57] The business is trying to become more efficient and you have to do the work to

[00:40:02] ask the deeper questions and understand what time you're spending as an analyst

[00:40:07] that is spent helping the business improve and may not be useful.

[00:40:12] And my, my job at Hamilton was a great example.

[00:40:15] I was working for the CEO of Hamilton reporting to him directly.

[00:40:19] And, um, I got bombarded with the most useful questions to the business all day,

[00:40:24] every day, because he was coming to me and saying, this is the most important thing.

[00:40:28] I need these numbers to make this decision.

[00:40:30] Can you get that to me?

[00:40:31] And sometimes on it as a, as an analyst, you might be disconnected from, from the

[00:40:37] decisions that are being made with your data and the decision makers.

[00:40:41] So the closer you can get to understanding what the context is of the business owner

[00:40:46] or the business case for your data, the more useful your data will be.

[00:40:51] And the most, the more useful you'll be as an analyst.

[00:40:54] You gotta, you gotta know the business.

[00:40:56] Yeah.

[00:40:57] Businesses care about whatever they care about.

[00:41:00] Usually dollar bills.

[00:41:01] A lot of times it could be people reached, it could be live saved, but

[00:41:04] whatever it is, you got to figure it out and you got to figure out how does data

[00:41:08] equal more dollar bills or more life saves?

[00:41:11] Yeah.

[00:41:11] They don't, they don't care if you're using like some new Python library.

[00:41:15] They don't care if you're using CTEs or sub queries.

[00:41:18] Those are all things you care about.

[00:41:20] Those are your tools.

[00:41:21] The real important thing is are you delivering useful information to

[00:41:26] help facilitate business decisions?

[00:41:29] Because like, like you mentioned earlier, you're making decisions for a

[00:41:34] multimillion dollar marketing campaign.

[00:41:36] If you do good with those decisions, you make them multimillion dollars.

[00:41:41] Right.

[00:41:42] And like for that, for that request, actually.

[00:41:44] So we had a data change between theatrical releases.

[00:41:49] So our last, our last movie, we, we changed some systems to

[00:41:53] start relying on a new data stream.

[00:41:55] And if I hadn't checked my assumptions first before starting that request,

[00:42:00] I would have missed out on a lot of important data coming from a new source.

[00:42:05] So really questioning and never really trusting like, Oh, some guy wrote this

[00:42:10] query, that's the one I'll use is, is how you can be most useful, I guess.

[00:42:17] It's, it's easier said than done, but I think you do a good job at that, Paul.

[00:42:21] I've always, one thing I've always known about Paul is he asked good questions.

[00:42:24] Actually, these are two things I know about Paul.

[00:42:26] He might do it too much to a fault.

[00:42:28] I don't even know.

[00:42:29] You ask a lot of questions and you got an action to bias or gosh,

[00:42:33] you got biased to action.

[00:42:34] Maybe you got action to bias too.

[00:42:36] I mean, maybe you're a biased guy.

[00:42:37] I'm so biased about action.

[00:42:39] I hope you're not biased, but like those are two things I think

[00:42:42] that you've always done well, Paul.

[00:42:43] And I think that's led to, you know, a lot of success for you transferring

[00:42:47] your career into data and doing well in that career.

[00:42:51] I think those two things have probably taken you a long way.

[00:42:54] Thanks.

[00:42:55] One last thing I want to bring up, I think a useful framework for understanding

[00:43:01] how to ask better questions, on how to ask better questions is

[00:43:06] the Toyota Five Wise methodology.

[00:43:08] Okay.

[00:43:08] I don't know it.

[00:43:08] Teach me.

[00:43:09] So it's, it's a lesson from manufacturing that you can use with your job.

[00:43:13] So one question might be in the context of, of the Toyota manufacturing,

[00:43:16] they would say this part broke and then they'd ask, why did that part break?

[00:43:19] Oh, it's because this machine bent to the wrong way.

[00:43:21] Why did that machine bended the wrong way?

[00:43:22] Well, it's because this guy didn't do the maintenance on that machine.

[00:43:25] Why didn't he do the maintenance on that machine?

[00:43:27] Well, it's because he was overscheduled and there weren't

[00:43:29] enough people on that team.

[00:43:31] Why are we, why was he overscheduled?

[00:43:32] Why are there not enough people?

[00:43:34] And you get to the, the, the core problem, three or four or five steps

[00:43:38] removed from the surface problem.

[00:43:41] Or the proximate cause, which is that, you know, some, something

[00:43:46] completely almost unrelated.

[00:43:48] And so I think if you're, if you're taking that kind of approach as well in,

[00:43:53] in your job, you can get to the right answer a lot faster.

[00:43:56] So like for, for example, we're talking a lot about data

[00:43:59] modeling at my company right now.

[00:44:00] So it's like, why is it so hard to answer a question about how much

[00:44:04] revenue we have for this very specific thing?

[00:44:06] Well, it's because of this.

[00:44:08] Well, why is it?

[00:44:08] Why, why is that the case?

[00:44:10] Well, it's because of this.

[00:44:11] Well, why is that the case?

[00:44:12] Why is that?

[00:44:12] And then you, you, you come five steps down until you get to a root cause.

[00:44:17] And then if you focus on solving that thing instead of the surface problem,

[00:44:21] you will be a lot more useful to the business and solve a lot more problems.

[00:44:25] Make more Buko bucks.

[00:44:27] And make more Buku bucks.

[00:44:28] Yeah.

[00:44:29] I love it.

[00:44:30] What was the five Y from Toyota?

[00:44:32] Toyota five wise.

[00:44:33] Yeah.

[00:44:33] There you go.

[00:44:34] Well, Paul, it's been great having you on the podcast.

[00:44:37] Thanks.

[00:44:38] I appreciate all of your thoughts.

[00:44:40] If people want to learn more about you or reach out to you, where should they go?

[00:44:44] I guess you can reach out to me on LinkedIn.

[00:44:46] Paul Alex, all-stream.

[00:44:48] You guys should check him out at least, at least give him a follow on LinkedIn.

[00:44:50] He's got some, he got some good SQL videos coming out that I saw the other day.

[00:44:54] Oh yeah.

[00:44:55] I couldn't sleep on Saturday, so I made some SQL training videos.

[00:44:58] If you want to take a look at those, I'm planning on going through all

[00:45:02] the practice problems on that site.

[00:45:04] Okay.

[00:45:04] Awesome.

[00:45:05] Thanks Paul.

[00:45:06] Yeah.

[00:45:11] Hey, I hope you guys enjoyed that episode.

[00:45:14] Thank you again to Paul for giving us all that great insight on how he transitioned

[00:45:19] from a non-technical background to a data analyst role and what he does in

[00:45:23] the day of a life as a data analyst.

[00:45:25] We appreciate it so much, Paul.

[00:45:27] Once again, just a quick plug.

[00:45:28] I want you guys to check out some of the resources down below in the show notes,

[00:45:31] specifically Average GPT.

[00:45:33] You guys got to check this out.

[00:45:34] It's super cool.

[00:45:35] It's free for right now.

[00:45:36] Maybe I'll start charging in the future.

[00:45:38] I'm not sure, but go use it right now while it's free.

[00:45:40] All right.

[00:45:40] Thank you guys for listening.

[00:45:42] Go listen to another episode.

[00:45:43] Why not?

[00:45:43] Right?

[00:45:44] Have a good one.

[00:45:44] Bye.