Reacting to "FASTEST Way to Become a Data Analyst & ' Get a Job" video by Stefanovic

0:00Hey, it's Tim here. In today's video, it's

0:02a completely different type of content.

0:04I've been approached by quite a few data

0:05analysts saying, "Hey, Tim, why do you not

0:08make more

0:09content about how to become a data analyst?

0:11How do you not talk more about your career

0:13and everything else?" And the main reason

0:15is because it's actually quite hard to do

0:17that.

0:17And I think a lot of journeys are also

0:19quite unique. And unless you talk to a lot

0:21of people,

0:21you won't actually understand the variance.

0:23But on YouTube, there has been this surge

0:25in what I'd

0:26call influencer data analysts who are

0:28essentially talking about their careers and

0:29their journeys

0:30and showcasing how they've got to where

0:32they are. Stefanovic is one of those people

0:35who's

0:35actually created a pretty big following and

0:37I think some really good videos around how

0:38to

0:39become a data analyst. I'd highly recommend

0:40actually you subscribe to his channel

0:42because

0:42he talks very openly and honestly about his

0:44experience. And that's actually very

0:46valuable

0:46because it's only by listening to people's

0:48experiences will you learn how you can

0:50improve

0:51and get better. That's also the point

0:53behind channels like mine. But what I want

0:55to do is

0:56react to a video that he made about the

0:58fastest way to become a data analyst and

1:01actually get a

1:02job. This is the most personal one because

1:03that's what pretty much everyone reaches

1:05out for.

1:06And it's had 1.5 million views, which is an

1:08incredible achievement in itself. To get to

1:10that kind of milestone on YouTube takes an

1:12incredible amount of work. You have no idea

1:15.

1:15So I want to watch this video and hopefully

1:17give you my reactions to it. So you can

1:19hopefully learn

1:20a bit about what I think of this, but also

1:21maybe also challenge some of the

1:23perceptions that you

1:24might be picking up incorrectly from this

1:26video. As ever, let's get stuck in. Okay,

1:29so let's get

1:31started. From the day I started messing

1:33around with some data in Excel, it took me

1:35about three

1:36years to land a data analyst job at Heine

1:38ken and another two years to land a job as a

1:40freelance data

1:41analyst at a big bank in Europe. And

1:43another two years to quit that job and

1:44travel the world full

1:46time, but more on that later. So that's

1:48pretty interesting that that's how he

1:49starts out. I

1:50think he starts out with I think it starts

1:53out with an unrealistic vision. You see,

1:56he like to do the kind of transition he's

2:00doing. I think you have to be exceptionally

2:03skilled. And I

2:04think you also have to be incredibly lucky

2:07with not just the economy, the market, but

2:10also the

2:10opportunities that land at your feet. You

2:13've almost got to learn a skill, progress to

2:15the

2:15next thing, almost immediately learn the

2:17next skill. And as you're sort of perfect

2:19ing that and

2:20refining that, by the time you get to sort

2:22of, you know, the ceiling and that another

2:25opportunity

2:25comes up, and then it perfectly lands in

2:27your plate, and then you take it and you

2:28get it and

2:29you go on. For many analysts, that's not

2:31the normal route. And so I think this is an

2:34inspirational

2:35message. And I think you should hear it as

2:37exactly that. But it can also be a little

2:39bit demoralizing

2:40if that's not the journey you've been on.

2:42So I always like to just preface these

2:44things. I mean,

2:45the title of the video is the fastest way

2:47to become a data analyst and actually get a

2:49job.

2:49And it's his method, it's his approach to

2:54this. So just always be careful with sort

2:59of these

2:59timelines that people give you. And of

3:01course, YouTube is designed to kind of pull

3:03you in. So

3:03he's got to lead with something innovative.

3:05He wants to inspire you. So that makes

3:07total sense.

3:08So let's carry on. You become a data

3:11analyst way faster. 100%. I've wasted too

3:13many hours watching

3:14YouTube tutorials, and I've spent money on

3:16Udemy courses that didn't add any value. So

3:19I often ask

3:20myself, that is absolutely true. The

3:22problem with Udemy is that they take your

3:25money. And you often

3:27think that once you've purchased the course

3:29, you've you have access to this knowledge,

3:30very

3:30few people actually finish the Udemy

3:32courses, and good pace and then apply those

3:35skills to real work

3:36to them actually capitalize on the value

3:38that was held inside of that course. So it

3:41is totally

3:41harmful for lots of people to be buying

3:43courses, and not seeing the end result.

3:45What you have to

3:46be doing is be incredibly motivated,

3:48incredibly driven, make sure you push

3:50through it as simple

3:51as that. If I could start over with data

3:53analysis, knowing what I know today, how

3:56would I go about it?

3:57And that's exactly what this video is about

4:00. Remove all the fluff and give you the

4:01fastest

4:02path to go from zero to a full time data

4:04analyst. And throughout this video, I will

4:07share the three

4:08major mistakes almost every data analyst

4:10makes. So stay tuned for those. The first

4:13thing every

4:13beginner data analyst needs to pick is a

4:15tool or programming language for data

4:18analysis. I

4:18personally started off with Excel. Is that

4:21the best tool there is for data analysis?

4:23No. If I

4:24could go back in time, would I have started

4:26with something else? Also no. Because

4:28apparently getting

4:29good at Excel was enough for my first data

4:31analysis job. And I dare, I hate to say I'm

4:34here as Tableau Tim. He's absolutely

4:37correct here. To get your first data

4:39analyst job,

4:40most people think you have to go learn

4:42Tableau, go learn Power BI, because those

4:44are the jobs that

4:44are in demand. But it's also just as

4:47effective to learn something like Excel. Go

4:51somewhere where

4:52they only have Excel, show your value there

4:55, then be the one to introduce them to new

4:58and better

4:58tools. That's also just as effective way.

5:01And you might find that something like

5:02Excel has a much

5:03bigger sort of library of resources you can

5:06pick up. It's pretty much almost accessible

5:09everywhere.

5:09And a lot of the skills you learn working

5:12with Excels are transferable, like a lot of

5:14the

5:14formulas, a lot of the techniques, you can

5:16then go and apply those in other

5:17technologies. It's all

5:19analytics at the end of the day. So it's

5:21not unique, necessary to Excel or other

5:23tools. So

5:24it's actually a really good way to start.

5:25The other reason I think Excel is actually

5:27super

5:28important is because a lot of organizations

5:30still use Excel alongside other platforms.

5:33And so what

5:34you'll find is you'll go in as a data

5:35analyst, and you'll have to connect to

5:37Excel, open in Excel,

5:39decipher what's in the Excel and understand

5:41how it works. If you have the ability to

5:43open an Excel

5:43file and understand everything going on in

5:45there without having to go back to people

5:47and ask for

5:47help, you have a superpower that you just

5:51don't appreciate. As annoying as it sort of

5:54pains me to

5:55say, knowing your Excel skills is super

5:57important. And I think if you're going to

5:59start somewhere,

6:00start with what you have and nearly

6:02everyone, I can't think of anyone who doesn

6:04't have Excel,

6:04if you don't have Excel, you probably have

6:06access to Google Sheets for free, just open

6:08up a Google

6:09account, go and learn the exact same skill

6:10sets there, build something, script

6:12something in Google

6:13Sheets, and then take that knowledge into

6:15Excel, almost all the features again, you

6:17know, pretty

6:17much one to one. So yeah, don't don't sit

6:19there and sort of come up with excuses

6:21about what you

6:22don't have access to, things are expensive.

6:24Just start with what you have. And as long

6:26as you make

6:26that journey and you start showcasing what

6:28you're building somewhere, it's a pretty

6:30good way to get

6:31involved. All right, let's carry on. Job.

6:33And I personally recommend every single

6:35data analyst

6:36to start off with Excel. It's easy to learn

6:38, used in almost every company, and it has a

6:40wide range

6:41of applications for data analysis. But this

6:43is also its pitfall. Although you can do a

6:46lot of

6:46different things in Excel, it's not

6:48designed specifically for data analysis,

6:51meaning it has

6:52its limitations. You want to work with a

6:54very large data set. Well, good luck not

6:56punching

6:56through your computer screen. Very true.

6:58Excel crashes for the 27th time.

7:00Excel is perfect for starting out. But once

7:03you've mastered Excel, it's time to move on

7:07to the more

7:07serious tools. And the most logical tool to

7:10learn next is SQL. Next to Excel, it is

7:13apparently the

7:13second most requested skill in data analyst

7:16job openings. Now, just because it's the

7:22second most

7:23requested skill, to me doesn't mean it's

7:25what you should go to next. And I think it

7:28actually

7:29really depends what you're trying to do in

7:31the context you're going into. The great

7:36thing about

7:36SQL is it's a generic skill. You always

7:38have it on your CV. People always value it,

7:40and it's going to

7:41be fantastic. But if you're starting from

7:44zero, going from something like Excel to

7:47SQL almost

7:48feels like a step backwards, because you're

7:50going from what is a semi visual tool that

7:52allows you to

7:52do some basic charts to almost what is not

7:55a programming language, but you're working

7:58in SQL,

7:59you're not necessarily going to be, you

8:01know, creating things that inspire you, you

8:04're just

8:04going to be looking at SQL statements, left

8:05, right and center, kind of learning all

8:07this terminology.

8:08And I always think that sometimes it's also

8:10important to, to start to build something

8:13to

8:14show yourself some sort of reward for the

8:15skills you're having. And if you're going

8:17from Excel to

8:18SQL, I kind of think you're doing it

8:20backwards. So I'd actually recommend go

8:22pick up a data

8:23visualization tool, whether it's something

8:26like D3, Tableau, Power BI, Python,

8:28whatever it is,

8:29just go pick up something visual and start

8:31applying the things you've learned in Excel

8:34to those tools and start building things.

8:36Because it's only when you start building

8:38things,

8:39you start to realize what problems you have

8:41in your data. And when you start to realize

8:44what

8:44those problems are, then guess what SQL

8:47starts to make a ton more sense, because

8:50you're discovering

8:50problems, you're understanding how you need

8:52to shape your data, because you're

8:53literally looking

8:54and visualizing your data. And then when

8:57you come to learn SQL, you're doing it with

8:59a purpose,

9:00because you're trying to recreate solutions

9:01to problems you've already come across.

9:03Whereas if you go and learn SQL, having

9:06just worked in Excel, whilst Excel might

9:08have given

9:08you some of those challenges, it won't give

9:10you anywhere near as enough challenges as

9:13using

9:13something visual where you want to maybe

9:15build a certain chart type and the data is

9:17in the wrong

9:17format. So I don't know what he's going to

9:20say next. But that to me would be sort of

9:22where I

9:22slightly disagree with this. But let's

9:24carry on and see what else he says. Huge

9:26thanks to Luke

9:27Baruse for actually scraping a lot of

9:29LinkedIn job openings. I'll stop here and

9:32say Luke is fantastic.

9:33Go follow his channel. Yeah, put a link in

9:35the description. Luke Bruce has probably

9:38some of the

9:38most neutral data analyst content on

9:41YouTube. And he's done a really sort of

9:44good analysis of roles

9:47in the industry. But also, he's also looked

9:50at a lot of certifications and what's

9:51actually useful

9:52for data analyst. I think his video on the

9:55Google data analyst certification is

9:57probably one of the

9:58best ones out there. And yeah, generally

10:00his content is also funny. And it's

10:02interesting.

10:02And he's just got a good vibe. Go follow

10:04the channel and check it out.

10:05- He used to come up with these facts. You

10:07're a legend, man. Thank you.

10:08- What up there, nerds?

10:09- And the reason SQL is so in demand is

10:11because it doesn't have the same

10:13limitations as Excel.

10:15Why is my accent suddenly changing to solid

10:18? I don't know why. With SQL, you can...

10:21- I can share some... That happens when you

10:24're recording videos. I think Zefanovich is

10:27using

10:28a teleprompter here. And when you get sort

10:30of focused into a teleprompter, your voice

10:33can start

10:33to do... You just start paying attention to

10:35your voice because you're trying to pay

10:36attention to

10:36the pacing and the reading. So yeah, that's

10:38totally normal.

10:39- You can extract, transform, and load very

10:42large datasets. But the best thing about it

10:44is that SQL

10:45has its own very easy to use programming

10:47language. It's a great tool to add to your

10:49skill set as a

10:50data analyst while also being a great

10:52stepping stone to some more serious

10:53programming.

10:54- Yeah.

10:54- Before we get to that, every data analyst

10:57needs a visualization tool. There are

10:59dozens of tools out

11:01there. But having worked for multiple

11:03companies and projects and having looked at

11:05probably

11:06hundreds of data analyst job openings, I've

11:08come to realize that the majority of

11:10companies are

11:10looking for people with experience with

11:13Tableau, Power BI, and QlikView. Here's the

11:15...

11:16- And if I just go back, can I get that

11:18freeze frame there? No? Let's just see if I

11:22can time it.

11:22- ...companies looking for people with

11:25experience with Tableau, Power BI, and Qlik

11:27View.

11:28- All right. Let's stop there. So... Oh,

11:32when was this video made? Six months ago.

11:33There are more tools than these three. And

11:36I think he's absolutely right. The most

11:39roles you'll see

11:40do ask for Tableau, Power BI, and QlikView.

11:43Obviously, as Tableau, I'm going to say,

11:45"Learn Tableau." That's obvious, right? But

11:47let me also be very honest with you. Wh

11:52ichever one

11:52you pick first, great, perfect. But if you

11:55're trying to be a data analyst, you're

11:56trying to

11:57get into a role, just know that it won't be

11:59the only tool you'll need to learn, right?

12:02So

12:02unless you're sensational at Tableau and

12:05you kind of give yourself a runway of three

12:08years in an

12:08organization in the setting that only uses

12:10Tableau, in those situations, you're pretty

12:13lucky. But if you want to be flexible, if

12:15you want to be sort of as applicable to as

12:17many people,

12:18then you're going to need to learn, I think

12:20, two at the very least. Now, you could go

12:23with Tableau

12:24and Power BI, you could go with Power BI

12:25and QlikView, whatever combinations you

12:27want.

12:27But I also recommend actually doing

12:29something a little bit rogue, which is

12:31choosing one of

12:32these tools alongside a tool that's not

12:35here. Because you see, some organizations

12:38use tools

12:39you've never heard of. And in order to

12:41appreciate those tools, I think it's really

12:43good to sort of

12:43diversify what you're learning. And these

12:47three tools kind of sit in the same space,

12:50business intelligence tools made for

12:52businesses with visual interfaces. They

12:54each have their

12:55strengths and weaknesses. If you work with

12:57Power BI's data modeling capabilities are a

12:59little bit

12:59more advanced than Tableau. If you're

13:01working with Tableau's data visualization

13:03and flexibility

13:04techniques are a little bit more advanced

13:06than some of these other two. But in

13:07essence,

13:08all these tools are converging to sort of

13:09one vision of how these things should work.

13:11So

13:12you should go and learn something rogue, go

13:14learn something like ggplot. And it's a

13:16really

13:16random choice. But working with something

13:19else like that will also teach you to

13:21appreciate some

13:22of the things that are super easy in these

13:25tools, and what's hard elsewhere. And it

13:27also means that

13:28you can go for these kind of very unique

13:30roles where people are looking for very

13:32specific things.

13:33And so just learning one tool to me is not

13:35enough. I've actually said this in a video

13:37I did

13:37talking about the best ways to learn Table

13:39au, I actually said call out the fact that

13:41look,

13:42just learning Tableau is not going to be

13:43enough, you're gonna have to learn more

13:44things. So

13:45couldn't agree more, learn multiple tools.

13:48And you can start off with one of these

13:50most commonly sort

13:51of selected tools. But I think it's also

13:53important to diversify and learn something

13:56different.

13:56Power BI is part of the Microsoft stack. So

13:59work smoothly with Excel and SharePoint.

14:01Besides that,

14:02it has a free version. And even the big

14:04versions are relatively budget friendly

14:07compared to the other BI tools. Next up is

14:09Tableau. Tableau is a BI tool with more

14:11extensive

14:12data visualization capabilities than Power

14:15BI. It seems to be just a bit more in

14:17demand in the

14:18job market than Power BI. It comes with a

14:20much higher price tag though, making it

14:22harder to

14:23learn if you're starting out by yourself.

14:25And then there's... I don't agree with that

14:27. And I know for

14:28a fact that Stefanovic knows this, but

14:30there's a product called Tableau Public,

14:32that is a free

14:33version. It's got all the same capabilities

14:35as Tableau desktop. I've actually already

14:37done a

14:37video on this on my channel, go check it

14:40out. And the other thing is, and now we're

14:43treading into

14:44Power BI versus Tableau territory. But if

14:47you look at the sticker price of a product

14:50and think that

14:50that's the only cost related to deploying

14:53that software, you'll quickly learn at

14:56least as an

14:56analytics consultant that that is the wrong

14:59approach. Because these technologies have

15:01other costs. And the concept you always

15:03have to sort of think about something

15:05called the total

15:06cost of ownership. And that value is

15:08different in every organization. It's not

15:11just enough to

15:11look at the sticker price. So yes, Tableau

15:13might look more expensive per user and the

15:15pricing might

15:16be set up differently. But you've also got

15:18to understand that Microsoft includes Power

15:21BI as

15:21part of wider enterprise deals. And so if

15:25you think about sort of Excel and Power BI

15:28and Microsoft,

15:29Microsoft is pretty much on every

15:30enterprise laptop. So it's much easier for

15:32Microsoft to

15:32sort of bundle in Power BI with some other

15:34piece of software. And that's what I think

15:36makes it

15:37cheaper. Because in essence, it's easy for

15:39Microsoft to throw in Power BI and a bunch

15:41of other features for free to help lock you

15:44into their platform, specifically Microsoft

15:46Azure. So

15:47be aware of that because Tableau doesn't

15:49have any of those sort of other big cloud

15:51platforms. It's

15:52been acquired by Salesforce. But Salesforce

15:54doesn't have a cloud offering that's

15:55equally as

15:56competitive. But if you're working in the

15:58CRM world, and you use Salesforce as your

16:00CRM,

16:00Tableau might start to be a bit more

16:02competitive versus Power BI. But I never

16:06like to go into cost

16:06comparisons, because I don't think I know

16:08enough from a sales perspective to know

16:10what the cost of

16:11ownership is for different organizations.

16:13And also, you've got to be aware that Table

16:15au changed

16:15the price depending on your scale. So

16:17whilst you see the prices on the website,

16:20that's never really

16:21the real price when you go to enterprise

16:23level. The real price, it tends to be sort

16:25of bundled

16:25with a bunch of other things, not including

16:27Tableau desktop as well, a bunch of other

16:29sort of capabilities. But from an

16:31individual perspective, if you want to

16:34learn Tableau,

16:35use Tableau Public, you can use that

16:36straight away for free in the browser,

16:38download it to your

16:39desktop. And it works pretty much the same

16:41way as a full desktop experience. The only

16:43thing you can

16:44connect to is databases. That is the only

16:46catch. But you can connect to flat files,

16:48which means if

16:49you're just learning, you can go to Kaggle,

16:50download a dataset, work with it, you can

16:52go to

16:52Microsoft Excel, download a dataset, work

16:54with it. If you've got data in Google She

16:56ets, go to Google

16:57Sheets, download that data or connect to it

16:59directly from Tableau Public. That also

17:02works.

17:02It's pretty flexible. And you can take part

17:04in all the sort of learning activities that

17:06are

17:06available in the community as well. So just

17:08a word of caution there when people talk

17:11about prices,

17:12I don't know enough about prices. And I'd

17:13like to think I know Tableau very well, and

17:15I don't feel

17:16comfortable talking about it. So something

17:17to watch out for. ClickView, using in-

17:20memory technology,

17:21making it a super fast and responsive way

17:23of doing business intelligence, but also a

17:27pretty

17:27high price tag. And also it's less in

17:30demand than Tableau and Power BI. I

17:32personally think...

17:32That's interesting. The in-memory thing,

17:34all the tools work in memory. I'm not sure

17:37why you

17:37called it out specifically. Tableau has an

17:39in-memory technology called Hyper. I'm sure

17:43Power BI has the same sort of technology.

17:45The reason that's important is because the

17:47in-memory

17:48capabilities is what allows these pieces of

17:50software to pull up and load data really

17:51quickly.

17:52It loads the data into memory rather than

17:53having to read it from the raw file every

17:55single time.

17:56Pick Power BI as my BI tool, although it

17:58might not be the fastest BI tool out there

18:01or the one with

18:02the most extensive capabilities. In my

18:04opinion, it can give you the most bang for

18:06buck when it

18:07comes to becoming a full-time data analyst.

18:09Before... Really interesting. The most bang

18:12for buck.

18:14What does that mean? I've seen people build

18:18incredible Tableau public visualizations

18:21and

18:21land jobs almost instantly as well. Because

18:24I've never used Power BI and maybe I should

18:27do one time,

18:28but I've never used Power BI, so I can't

18:29speak to that. I'd love to know what your

18:31thoughts are

18:32as a data analyst. If you've learned both

18:35of them, which one of the assets you've

18:37created in either

18:38software actually kind of has the most

18:40appeal when you put it in your portfolio or

18:43your CV,

18:43that'd be a really interesting question to

18:45know or answer to that question.

18:47Before we move on to how to actually get

18:50good at data analysis and land that first

18:52job,

18:53there's one more thing we need to do as a

18:54data analyst. After you've mastered a BI

18:57tool,

18:57it's time to get into the more advanced

18:59analytics. I'm talking about which

19:01programming language to

19:02use when it comes to data analysis for data

19:05science. Software developers and

19:07programmers

19:07get to choose between Java, JavaScript,

19:11Ruby, C#, Python, R, C, C++ and many more.

19:16Luckily for us data analysts, we only have

19:17to choose between two programming languages

19:20,

19:20R and Python. So which one should you pick?

19:23I chose Python, but if I could go back in

19:26time,

19:26I would choose Python. Why? Well, even

19:29though they're actually very similar, R is

19:31a programming

19:32language focused more on statistical

19:34analysis, while Python is a more general

19:36purpose programming

19:38language that also happens to be very good

19:40for data analysis. And although R might be

19:43a little

19:43bit easier to learn, I would still go for

19:47Python as it's probably the number one

19:50programming

19:51language in the world. Meaning if you get

19:53good at Python for data analysis, it might

19:55also open

19:55the door to a lot of different job

19:58opportunities. So that's an interesting

20:01take. I was waiting to

20:02see where he went with it because I think

20:04if you learn if you learn SQL as he

20:09suggested,

20:10if you learn a visualization tool as he

20:12suggested and you'd learn Excel and then

20:15you'd learn two

20:16other you know tools, you're already pretty

20:19much good to go. I think in most enterprise

20:22context,

20:23you don't need R and Python. R and Python

20:26tends to be reserved for specifically data

20:29science roles,

20:30and this is a data analyst role. So data

20:33science roles require R and Python by

20:36default, but as a

20:37data analyst, as an entry level data

20:40analyst, I don't think you need R and

20:42Python out of the gate.

20:44It doesn't hurt to learn it and it doesn't

20:46hurt to be aware of what it can do and it

20:48doesn't hurt

20:48to spend let's say a weekend on that Udemy

20:50course. I said you probably won't finish.

20:53Just spending

20:53four hours going through Udemy course

20:55building something, putting it on GitHub so

20:56you can show

20:57people that you've actually least dabbled

21:00with it. But do you need to be sitting

21:02there writing R and

21:04Python scripts to solve data problems out

21:06of the gate? I'm not convinced. Might you

21:09need to do that

21:10in the future for your career progression?

21:12I think so, especially if you want to

21:13become a data

21:14scientist or go down that route. So you

21:16know, take it or leave it, but I'm not

21:18convinced you need R

21:20and Python. The path I took is Excel, SQL,

21:24Power BI and Python. This is the path...

21:26Interesting.

21:27That's really interesting. So and this is

21:30actually what the most watched segment in

21:32this video,

21:32right? So I'm not sure. So the path I took

21:38is Excel, SQL, Power BI and Python. Right,

21:44so let me give you like an honest view on

21:47these skills, right? So I'm a Tableau

21:51visionary. I've

21:52worked with Tableau for over seven years.

21:54So if I replace Power BI with Tableau, if

21:57you ask me,

21:57do I have Python skills? No. Do I have SQL

22:00skills? Yes. I have, you know, okay SQL

22:03skills. They're

22:04not, you know, they're not tier and S tier

22:06level SQL skills, but I can open up a SQL

22:09prompt,

22:10write some SQL, do some basic stuff. But

22:12the majority of my time is spent in Tableau

22:14.

22:14And also I use other tools, ETL tools like

22:17Alteryx, Tableau Prep, all those tools that

22:20some companies prefer over SQL. That's

22:22something too important to understand. We

22:23'll come back to

22:24that maybe later. And then of course Excel,

22:26just basic Excel. So I'm not an Excel ninja

22:28. I'm not

22:29writing VBA scripts, but I know my way

22:32around in Excel. So that's sort of my sort

22:35of place. And

22:35I've been working in the industry now for

22:37over eight years. I've worked as a

22:39consultant and a

22:39consultancy. I've worked at places like Acc

22:42enture, worked out in the field, leading

22:44teams. I haven't

22:46needed to push on, let's say SQL and Python

22:49to do that piece of work. Tableau has been

22:53the piece of

22:53software that has got me places into sort

22:55of roles. And in this case, that would be

22:57Power BI

22:58in Stefanovich's sort of role. So the skew

23:02here is that whatever tool presents the

23:06data tends to have

23:07sort of the biggest airtime in most sort of

23:10job applications. But when you actually get

23:13down to

23:13the work, you need data prep skills. And

23:15you could be using SQL or Python. But I

23:17think more companies

23:19are preferring tools that are a little bit

23:21easy to understand with a lower sort of

23:23adoption hurdle.

23:24And those tend to be visual tools, visual

23:26ETL tools to help you clean data. I've been

23:29even

23:29aware of data warehouse teams who've moved

23:31entire data warehouses without having to

23:33write SQL because

23:34they're using IDEs or instances that sort

23:37of allow them to do those transformations

23:39without having to

23:40write SQL. So just be aware that although

23:43this is his path, you know, go off the

23:46beaten path. You

23:47can make your own path and do it

23:48differently and hopefully make a YouTube

23:50channel and talk about

23:51it. Let's carry on. This is the path I

23:54would recommend as it worked out great for

23:56me. But

23:57you can start off with any language or tool

23:59that suits you best. Because whatever you

24:02pick,

24:02your first language or tool will definitely

24:05not be your last. Now that you know which

24:07tools to learn,

24:08let's talk about how to actually learn it.

24:10And this is where beginners make the first

24:12major

24:13mistake. The mistake most beginners make is

24:15that they try to learn by watching others.

24:18This is how

24:18most people learn. Some people go to Udemy

24:21and watch multiple 20 plus hour courses and

24:25probably

24:25not even finish all of them. Or they watch

24:27YouTube videos watching other people

24:29analyze data. But

24:30without actually writing code or analyzing

24:32data themselves, they give themselves a

24:34false sense of

24:35progress. Fantastic point. Amen. Amen to

24:39this. I was wondering where I was going

24:43with this.

24:44Because I was about to say no, absolutely,

24:47you should watch people and learn from

24:49those

24:49techniques. But you'll notice that I don't

24:52make videos where I go through sort of

24:54building

24:54dashboards. And I don't make videos where I

24:56go through try and build solutions. Instead

24:57,

24:57I make videos on features and show you how

25:00to use them. Because I very much agree with

25:02this sort of

25:02way of thinking. And you need to get hands

25:05on with the data yourself, right, you need

25:08to actually have

25:08a problem you want to solve and actually go

25:10through the process of doing it. Because

25:13only when you do

25:13it, do you hit the hurdles and make the

25:15mistakes that lead you to learn. When you

25:18're watching

25:18something like a Udemy course, or let's say

25:20really well curated YouTube video. The

25:23problem is, is

25:23those examples are set up to allow the

25:25formats to work. Let's say take Udemy, like

25:28you wouldn't want

25:30to watch a 30 minute Udemy video, someone

25:32troubleshooting something, it's not very

25:33engaging,

25:34you'd never make it through, you wouldn't

25:35finish the video. And you wouldn't want to

25:37watch a 15

25:38minute video, someone troubleshooting a

25:40problem inside of Power BI. Again, you just

25:43want to know

25:43what's the best way of doing it. But it's

25:45actually only by watching those troubles

25:48hooting methods and

25:49actually learning those methods, do you

25:51become a better person. And you're not

25:52going to watch those

25:53videos because they don't exist that just

25:55that I just don't perform well. People want

25:57answers to

25:57these questions rather than understanding

26:00where the questions come from. So go ahead

26:02and actually

26:03get stuck in with the data. I couldn't say

26:05this couldn't agree with this more. It's

26:06absolutely

26:07fantastic advice. I can't tell you how many

26:09people watch something they said they've

26:12watched

26:13something or read something and then you

26:14ask them sort of a follow on question from

26:15that and they've

26:16never thought of it because they've just

26:18not engaged with the content at a sort of a

26:20thought

26:20level. So really important. Because

26:22analyzing data in your head is very

26:25different from actually doing

26:27it. Yeah. Data analysis in your head is

26:31very different from actually analyzing data

26:35, stumbling

26:36upon faulty data and debugging for hours.

26:39So what is the right way to do it? The

26:41answer is very

26:42simple. Learn by doing. You just need to

26:44get the reps in and start coding and

26:46analyzing data by

26:47yourself. And no one wants to hear this. I

26:50feel like this is where I kind of, you know

26:54, the title

26:54and the message are kind of a little bit

26:56disconnected. Here's the fastest way to

26:58become

26:59a data analyst to actually get a job. By

27:00the way, it involves a ton of work. You

27:03just need to spend

27:04time. It's true. You're not going to get

27:07good unless you spend time doing it. Some

27:10of the most

27:10talented people in the Tableau world have

27:12just simply spent time working on skills,

27:15building,

27:15building, building, building. And you can

27:17see their progression through their Tableau

27:19public

27:19profiles. You can see their progression

27:21through the way they approach work. And you

27:23only get that

27:24by getting paper cuts along the way on the

27:26journey and then learning from those

27:28mistakes. So if you

27:30can find places and resources that gives

27:32you challenges, absolutely go and get stuck

27:34in. I

27:34think that's what he's about to cover here.

27:36In the Tableau world, we have lots of

27:38different sort

27:39of community projects that allow you to hop

27:41into challenges every single week. I'll put

27:43a link to

27:43them on screen. I'll put it maybe an

27:45overlay right now so you can see where that

27:47is on the Tableau

27:48website. But nonetheless, I couldn't agree

27:50with this more. It's so, so important. I

27:52would recommend

27:53a website called ExcelPracticeOnline.com, a

27:56free website where you learn to use Excel

27:59by doing

27:59from the basics to the most advanced

28:01functionalities. For SQL, I would recommend

28:04the website w3schools.com/SQL. For Power BI

28:07, I would recommend DataCamp. There's free

28:10courses

28:11and also paid ones. For Python, you can go

28:13to this free website called learnpython.org

28:16. Whether

28:17you're practicing every day at your current

28:19job so you can make that switch to a data

28:21analysis career

28:22or whether you're still studying, you need

28:24to practice every single day. Only this way

28:26will you

28:27experience what it's actually like working

28:29as a data analyst. That's really good. And

28:32he's

28:33covered a couple of free resources there. I

28:35'm trying to think what's the Tableau free

28:36Tableau

28:37learning equivalent? It's got to be Tableau

28:39Public. And I think it's a combination of

28:41Tableau Public

28:41and the weekly challenges. Tableau has like

28:44the has some resources that you can get

28:46stuck in with.

28:47But what is interesting, maybe someone

28:50should make like a curated list of almost

28:52like challenges

28:53that you can go through in a specific order

28:56that teaches you skills in a certain way,

28:58just by repurposing the content that's

28:59already out there. Maybe that's an

29:01interesting idea. Maybe we

29:02should try that. Analysis. Working as a

29:04data analyst. Only then will you learn what

29:06it's like

29:07to debug your own code or formulas. Because

29:09to be honest, that's what you will spend

29:12the most time

29:12on. Debugging faulty code or correcting

29:15faulty data, being a data analyst is

29:18awesome. But this

29:20is also the second major mistake most

29:22beginners make. They try to solve every

29:24problem themselves.

29:26The fact is that error you just got in

29:28Power BI, there's a very big chance someone

29:30already has

29:31experienced exactly the same error as you.

29:33And that's good news for you, because now

29:35you don't

29:36have to look for the cause of the error.

29:37You just have to copy part of the error

29:39message in Google

29:41and you probably end up on some Stack Over

29:43flow site where someone explains exactly how

29:45to solve

29:46this error. Let's be honest, copying code

29:49from Stack Overflow is like 80% of any

29:51developer,

29:52programmer or data analyst's job. Data

29:55analyst is awesome. Okay, now that you know

29:57what tool...

29:58It's just funny. This drives me nuts. The

30:02amount of times I will get a question on a

30:05video or a

30:06question from someone in the community and

30:10their question is so precise that I'm able

30:13to take that

30:13question, put it into Google, literally hit

30:18enter. And the first link is the solution

30:22to said

30:22question. It drives me nuts, but people don

30:26't do it enough. And here's another thing.

30:29Unless you

30:31do this often, Google won't actually get

30:33good at helping you when you get stuck. So

30:36I've actually

30:37seen instances where when I Google for

30:39things, things that I'm looking for come up

30:42more often

30:43and the same resources don't come up for

30:45other people Googling the same thing. Why?

30:48Because I've

30:48used Google so much that it's just learned

30:51sort of the subjects and domain and the

30:53kind of things

30:53I'm looking for, that it nearly always

30:55pulls up the right answer in the first two

30:57or three hits.

30:58And unless you do the same thing and start

31:00teaching Google, "Hey, I like to search

31:02about

31:02data analyst things about these particular

31:04technologies and these particular solutions

31:06,"

31:06you won't teach Google what you're kind of

31:08really looking for. And Google does have

31:10some sort of

31:11mental model for each and every one of us

31:13over time. And in order to sort of build

31:15that up,

31:16you've got to start using it alongside your

31:18skill. So I nearly always work in a sort of

31:20dual screen

31:20setup. If I don't have that, I have a

31:22tablet or phone, I'll just quickly Google

31:23something on my

31:24phone and go out and try and sort of the

31:26answer in Tableau itself. Now, the other

31:29thing I'll say is

31:30that like in this particular case, he's

31:32copying solutions from Stack Overflow.

31:34Stack Overflow

31:35and Tableau isn't quite as common, but

31:37there is something called the Tableau

31:39Community Forums.

31:40With that, what I always say is sometimes

31:42people don't ask good questions,

31:45or they ask a good question, but it's

31:47really hard to replicate the problem unless

31:49you give people

31:50a sample data set or solution. So I always

31:53recommend go to Mockaroo, build a mock data

31:56set

31:56that allows someone else to try and solve

31:58the problem. Because if you do that, A, you

32:00engage

32:00more people, they actually want to try and

32:02come and help you. And second, it teaches

32:04you to better

32:05articulate what you actually need help with

32:07. That's another key thing that people don't

32:09often

32:09understand. Asking good questions leads to

32:12better answers. Ask bad questions, no one

32:15will answer you

32:16at number two. When you do get an answer,

32:18it will probably not be the solution. So

32:20learning how to

32:20ask good questions is also just as

32:22important. - A programming language to pick

32:26and how to actually

32:27learn it, it's time to build your portfolio

32:29. And this will help you build out an

32:31attractive CV

32:32for recruiters on LinkedIn. Recruiters, I

32:34always have talked like a black dish, I'm

32:36not a programmer.

32:38- So I'll use that sort of a little comedic

32:40moment to stop and say this. Not enough

32:43people

32:43think about building their portfolio as

32:45they're building up work. I almost think

32:47like everything

32:47you should you do, everything you do, you

32:49got to think of some way of adding it to

32:50your portfolio.

32:51And you might think, I can't add everything

32:53to my portfolio. And you're correct. But

32:55the thing is,

32:55it's only as you add things, do you then

32:57start to get a sense of prioritization of

33:00what's valuable

33:01and what's not. Because as you're doing

33:02things, as you're achieving things, you

33:04have a sense of sort

33:05of achievement firstly, but also, you'll

33:07start to realize, look, these are the

33:09solutions that really

33:10sort of created huge impacts. And these are

33:12the solutions that you know, actually maybe

33:14didn't,

33:14and I thought they did. The other thing is

33:16also teaches you to better articulate the

33:18impact and

33:19value you present to any organization. And

33:21in many ways, that is also a sort of

33:24another skill you

33:25have to learn because building all these

33:27things in Excel, Power BI, Tableau, to you,

33:30they're almost

33:31like levels in a game. And as you achieve

33:33those levels, you clear the levels, you

33:35beat the boss,

33:36great. But in business context, actually

33:38describing the value those things generate,

33:40it's actually quite hard. So if you start

33:42sort of doing this almost reflective

33:44approach of

33:45every time you do something, save it

33:47somewhere, put it in like a note, and then

33:49you understand

33:50over time, the kind of things and kind of

33:52problems you're solving, you can build a

33:54high level picture.

33:55And it's the high level picture. It's

33:57actually super valuable. For example, I

33:59could build lots

33:59of visualizations for fast moving consumer

34:01goods companies, fast moving consumer goods

34:04is

34:04essentially just, you know, people that

34:05make biscuits and chocolates and that kind

34:07of stuff,

34:07crisps, things that you know, you buy in a

34:09supermarket. Now I can build lots of data

34:12visualizations in that space. But as you

34:15start to work on more of them, you tend to

34:17realize that they

34:18generally fall around certain high level

34:21topics, for example, promotional spend,

34:24trade spend,

34:25optimizing promotions, these are the high

34:27level terms that you know, people in

34:29specific sectors

34:30are actually looking for. And those are

34:31what you should be using when you talk

34:33about your skills

34:34and your portfolio. So here's a data

34:36visualization, they blah, blah, blah,

34:39having the same sort of

34:39moment, here's a data visualization, it

34:42talks about promotions, I use it to help

34:45improve the

34:46trade spend for a particular organization.

34:48Okay, that's when you're starting to really

34:51connect the

34:52business sort of area and the data

34:54visualization, the actual work you're doing

34:56. No one's really

34:57interested in the end product, they're

34:59interested in the impact it has in

35:01businesses and in real

35:02work. And it's actually if you're good at

35:04articulating that, you'll find it very easy

35:06to talk to recruiters and extremely easy to

35:08talk about your accomplishments when you're

35:10getting

35:10recruited or hired. programming language

35:13and recruiters talk to me every day. Now if

35:16you have

35:17a job where you can apply data analytics

35:19already, then do so right away. If you don

35:21't really do

35:22anything data analytics related, like sales

35:24, for example, then try to build a report or

35:27a dashboard

35:28showing some key metrics or KPIs that you

35:30can show to your manager, he will

35:32appreciate it and you

35:33will get to learn your data analysis skills

35:35. And if you can, and you can go broader

35:37than this,

35:38you don't have to just do something work

35:40related, you can also go find public data

35:42on Kaggle or

35:43wherever you find your data sets. There's

35:44just tons of places to find data sets these

35:46days.

35:47Take a political issue today, take a global

35:49issue today, like climate change, take any

35:51of these

35:52issues, take these data sets, and tell a

35:54story, put them in your portfolio. People

35:56don't really

35:57care about the topics they care about

35:58seeing your skills. And it's only by trying

36:00different data sets

36:01and different skills, do you actually come

36:03across challenges that come across when you

36:05're just

36:05working with data sets, I've sometimes

36:07connected to data sets and found there's

36:08very little to tell.

36:09And I've actually realized that there's a

36:11plenty of story to tell, I'm just not

36:13seasoned enough

36:14at working with this kind of data to tell

36:16stories. And so by doing lots of different

36:18challenges,

36:19you'll actually learn a lot more about how

36:21to tell stories with data. In fact, there's

36:22a really good

36:23book called storytelling with data, I'll

36:25put an image of it and link to it here in

36:27the screen.

36:28It's super valuable. I think you should

36:30almost definitely read this. In fact, Luke

36:32Bruce has

36:32done a video on this already, summarizing

36:34the book. So go ahead and check it out,

36:36just buy the

36:37book. It's just you just have to have on

36:38your shelf, I have it somewhere and my

36:40books aren't

36:40on my shelf behind me. But just go get the

36:42book. It will teach you so much about data

36:45storytelling

36:46and analytics and being an analyst in

36:47general. Can't apply to your current job?

36:51Or if you're

36:51still studying, then start a project on

36:54your own at home. Because in the end, it's

36:56about getting

36:57that data analyst job. And the best way to

36:59do that is as follows. Pimp up your

37:01LinkedIn profile.

37:03I've made a video before where I gave my

37:05top five LinkedIn profile tips to attract

37:07recruiters on

37:08LinkedIn. And I'll summarize it here for

37:10you. Make sure you put data analyst in as

37:12many places as you

37:13can in your profile. LinkedIn is a search

37:15engine. And if recruiters search for data

37:17analyst, you

37:18want to be found, you want to be on top of

37:20the list. So make sure your header contains

37:22the word.

37:22If you're looking to get hired, Pimp up

37:24your profile. It's just very

37:25straightforward.

37:26Put in the keyword data analyst. Once your

37:28LinkedIn profile is set up, there's two

37:29things you

37:30can do. Start applying to every single data

37:32analyst job you can find on LinkedIn or any

37:35other

37:35job platform. Or the second way, which I

37:37prefer, is if you've set up your LinkedIn

37:39profile correctly,

37:40then recruiters will soon start hunting you

37:43. They will invite you to job interviews for

37:45data analyst

37:46positions. Which brings me to major mistake

37:48number three. Don't quit. Well, do quit.

37:53Let me explain.

37:54You see, the second best decision I have

37:56ever made in my life was becoming a data

37:59analyst. And to

38:00become one yourself, the only thing you

38:02have to do is don't quit. Whether it will

38:04take you one week,

38:06one month or one year. If you keep applying

38:08, if you keep improving your skills,

38:11updating your CV,

38:12you will land that job eventually. And

38:14trust. And this is such an important

38:16concept. In gaming,

38:17there's this idea called the grind.

38:20Essentially, when you're playing a game,

38:22they build in

38:23mechanics that just make you spend time in

38:25the game. You don't really enjoy doing them

38:27, but you

38:27have to go and do them to maybe gather

38:28resources, you know, upgrade something. And

38:30that's typically

38:31known as the grind. Now, it doesn't sound

38:34pretty nice. But when you work in the data

38:36analyst,

38:37there is an element of this, you do have to

38:38sort of grind through challenges and work

38:40through things

38:41to gain experience. And it just takes time.

38:43And some people almost quit a little bit

38:45early,

38:46because I think we come from a world where

38:48you're fed answers very easily and very

38:50quickly. And so

38:51people just expect to be told what hoops

38:53they need to jump to, to kind of get to the

38:55end result.

38:56And the reality is, is that if you jump the

38:58same hoops as someone else, well, how does

39:00that

39:00differentiate you? It doesn't. So you've

39:02got to go out and find your own challenges,

39:04tell your own

39:05story, and build up your own sort of

39:06combination of skills that are unique to

39:09you that you can use

39:10to sort of tell a story better than anyone

39:12else. And that takes time. If you think

39:14about it,

39:15one and a half million people watch this

39:17video. So if they all go and do what Stefan

39:19ovich has done,

39:20even if only 2% of them actually are

39:22successful, that's still thousands of

39:24people,

39:25they're going to have the same level of

39:27skill sets as you. So how do you

39:28differentiate yourself

39:30from those people? Well, you got to do a

39:32little bit more and you got to keep going.

39:34And it's

39:34exactly what he's saying here, don't quit,

39:36keep going. Many of you people don't make

39:38it to the end.

39:39But even if you make it to the end, you've

39:41still got to keep going and try and divers

39:43ify yourself

39:43in some way or form. Because there's just

39:46so much demand for people. And the specific

39:48thing that

39:49people are looking for, whether it's skill

39:51set, a particular tool, particular

39:52capability is actually

39:54soft skills. And soft skills isn't

39:55something that he's touched on in this

39:57video. But it's actually

39:58super, super important. I'll come to it

40:00very, very soon. Let's carry on. Once you

40:03do land that first

40:04data analyst job, a world of opportunities

40:06will open up for you. You can grow your

40:09skills into a

40:10senior data analyst or become a freelance

40:12data analyst, or what I eventually did,

40:15quit my job as

40:16a data analyst. The single best decision in

40:19my life was quitting my job as a data

40:21analyst. There

40:22is some sort of beautiful sort of things

40:25this someone became a data analyst in three

40:28years made

40:28a video about it, it got one and a half

40:30million views, and he actually doesn't work

40:32as data

40:33analyst anymore. Wonderful irony. See,

40:36becoming a freelance data analyst gave me

40:39the resources to

40:40quit my job and pursue a completely

40:42different career, creating my own brand

40:45while traveling the

40:46world. Do not subscribe to my YouTube

40:48channel. I'm 100% serious. If you're

40:50looking for advice and tips

40:52on how to become a data analyst and how to

40:54build a data analyst career for the rest of

40:56your life,

40:56then I would recommend subscribing to Luke

40:58Baruse. What up there, nerds? As he makes

41:00way better

41:01videos than me. But if you view being a

41:04data analyst as a stepping stone to

41:06creating a life

41:07you want, your dream life, whatever that

41:09might be for you, then you might want to

41:11subscribe as I

41:12share my personal lessons on his journey

41:15from a nine to five to a dream life. Watch

41:17this video

41:18next to learn what made me quit my six

41:20figure job to start YouTube. So that ending

41:24was pretty much

41:26almost what I see as a as a YouTube. I can

41:28kind of see what he's doing. He's pivoting

41:30his YouTube

41:30channel away from being data on this

41:32content. That's kind of what's enabled him

41:34to go and do

41:35what he's actually wants to do and enjoy.

41:37If I go back to his channel and let's just

41:39see his videos,

41:40they see 100%. Yeah, you can see that he's

41:43pivoted from let's see, when was this data

41:47analyst video?

41:47So 170,000 that lied nine months ago, this

41:51video is six months ago. So where is that

41:54video? So here

41:55you go. So fastest way to become a data

41:57analyst and actually get a job. That to me

41:59was the video

42:00where he pivoted and he's sort of pointing

42:02people towards Luke. All the videos after

42:04that are more

42:06about sort of professional lifestyle, he's

42:09kind of changing the channel and he's

42:11taking his audience

42:12with him because again, a lot of them will

42:14have that same aspiration. You want to

42:15become a data

42:16analyst to enable some sort of other

42:18ambition or desire. And you're seeing it as

42:20a quick way of

42:21doing that. So you know, hats off to Stefan

42:23ovich, he kind of cleared the bus at the end

42:25of this

42:26level. And he's off to do his thing. And

42:28yeah, definitely subscribe to Stefanovich

42:30if you want

42:31to go back and watch some of his older data

42:33analyst content. It's sort of interesting,

42:35because even if he does go off and do

42:37something else, I guarantee you he'll be

42:39using his data

42:40analysis skills to improve his capabilities

42:42and improve where he ends up. So once you

42:45're data

42:45analyst, you're always a data analyst. One

42:47thing I didn't see in this video is what I

42:50would call

42:51a focus on skills, soft skills in

42:53particular, you see, Stefanovich strikes me

42:56as someone who

42:57probably has a lot of soft skills. And it's

43:01hard to realize that because you've not

43:04seen him work

43:05in the settings he's working. He's also got

43:07enough soft skills to be going out and

43:08working for

43:09himself. He's built a brand, he's obviously

43:10got a great YouTube channel, which means he

43:12's a good

43:12communicator, whether he thinks that or not

43:15. And there's also a bunch of other things

43:17that you know,

43:18that there's so many things that make an

43:20individual and some of them what you see on

43:22YouTube

43:22is only but a fraction of what's actually

43:24going on. And so what I encourage you to do

43:27is follow

43:27lots of other data analysts, talk to people

43:29talk to them about their experiences and

43:31try and

43:32understand what soft skills have they built

43:35to kind of enable them to look at these

43:37challenges

43:38and almost brush them aside as they sort of

43:40go on and learn. It's a really, really

43:42vital thing.

43:43Maybe I'll do a video about what are the

43:45sort of top soft skills I think you should

43:47go with

43:47any particular data analyst role. I've

43:49already got a video that talks about how to

43:51learn Tableau and

43:52it does cover soft skills as part of that.

43:54And I go into a little bit of detail about

43:56business

43:56domain information as well. So go and check

43:59that out. But as ever, that's pretty much

44:01it for me.

44:02Hopefully you've enjoyed this little

44:03breakdown and I'll catch you in the next

44:05one.

44:05Transcribed by https://otter.ai

44:07Transcribed by https://otter.ai

44:11[ Silence ]