is data analyst a meme?

>>1721211
No they are paid pretty well. Big data is still a new and very much growing field.

Literally everything is a meme

shitloads of them in govt
they don't really do much, easy, high paid job, but difficult to get.

>>1721211
data analysts at large companies get paid shit loads. All my engineering tutors keep shilling the fk out of that job, like it's everyones dream. it does seem pretty hard to get though

>>1721211
What do they even do that is so difficult?

>>1721211

I'm in a Business Data Analytics B.S. at my business school.

It's bretty interesting and seems to have high utility.

>>1721224

if it has a buzzword like "big data" you can be sure its a meme

>company falls for the big data meme
>hires me as an intelligence analyst
>no one asks for anything
>f5 4chan all day

File: B1deRWuCMAANFhR.jpg (42KB, 600x600px) Image search: [Google]

42KB, 600x600px

>>1722084
>they don't really do much, easy, high paid job, but difficult to get.
Lol. Exactly. Hot girls get those jobs.

>look like this or gtfo

>>1721211

If you get hired, it's not a meme.

Otherwise, yeah it's basically a meme pipe dream.

>>1721211
I'm a "Business Analyst" on paper but essentially I'm a "Data Analyst". About a month shy of 10 years on the job.

I make about 72k in the midwest US.

My peers make anywhere from 60k to 110k. I could make more but don't want the headache. I've worked myself into a very niche role where everybody loves the shit I produce and nobody fucks with me or questions the value I add. I also have the ability to work from home (which I'm doing right now).

I deal mostly with Excel, BOBJ, SQL Server, SSIS and Teradata. A lot of pressure to use Tableau but I have literally zero reason to use it.

Shit's cash.

>>1725367
>BOBJ

I'm sorry you guys have to use SAP.

Kudos to you though. I'll be graduating with a BBA in Management Information Systems soon and imagine I'll go into an analyst position. Got any advice or tips?

>>1724141
This was my plan...
Seething with jealousy

>>1725367

Did you know most of those softwares prior to getting the job, or did you learn them on the job?

>>1725677
>I'm sorry you guys have to use SAP.

Agreed. I've heard we're going to be moving off of it at some point... which zero fucks will be given.

>>1725761
Learned on the job. I didn't start out as an analyst. I started as a peon and quickly learned the business. I had an itch to figure out why things were broken and/or how things could be improved. I would sift through data and figure shit out. Management loved it and the rest is history.

I know when people interview for our analyst roles off of the street, they must have solid analyst credentials to even be considered.

We simply wouldn't hire someone externally who doesn't know the inner workings of the business and doesn't already know how SQL works.

File: 1483521927742.jpg (7KB, 480x142px) Image search: [Google]

7KB, 480x142px

I'm a B.I. Analyst but that just means Data Analyst nowadays.

It's not a meme, and it's not even that hard to be honest, but if you have autism (protip: you do) it can be hard since most of the time you need to deal with directors, vice-presidents, CEOs, etc.

>SQL, Excel, Pentaho Data Integration (for ETL and data cleaning), Tableau, NoSQL

That's all you need.

>>1725891
>>1725848
Our school has taught us all a good deal about Excel, Access, several BI tools like Tableau, some SQL, the CASE tool Visual Analyst and lots of database theory and design. I'm enjoying it and seems very valuable. I've made it a point in my internship to learn SQL in my downtime and watch the guy I'm under do his reporting. Funnily enough he's a System Admin but they are a small firm and he fell into reporting, mostly because no one else there can do it.

If you two have any stories or advice I'd be very thankful.

File: 697.jpg (62KB, 399x399px) Image search: [Google]

62KB, 399x399px

>>1726083
The most important thing as a Data Analyst, in my opinion, is learning how to "clean" and "merge" data sources. Any small company nowadays have multiple data sources like Excel, MySQL, SQL Server, Oracle, Google Analytics, whatever. And the CEOs want what? Reports. Consolidated reports. They want dashboards where they can have a hard on just looking for that integrated data.

So learn ETL. Learn some data cleaning tool (Pentaho Data Integration is free, open source and widely used), learn how to merge databases and use the result as something you would apply Tableau to. This integrated result set is golden and if you make some nice graphs with it people will cum gold in your pockets.

>>1726121
ETL is one thing we haven't really covered at all yet, still a few more classes left in my major so maybe that will cover it, but I'll be sure to start playing around with Pentaho.

And from my current internship I've learned management loves graphs. They are barely a year into a new ERP system and still trying to get their bearings straight on reporting. All our meetings eventually talk about reporting issues.

>>1726140
I don't think they will talk about ETL, no one does. "Data cleaning" means that the data won't be good to be consumed and usually schools/universities will work with "good" data sources, avoiding almost completely the need to clean it beforehand. Real life is different, though.

Gotta love those "Country" columns with values such as:
>USA
>U.S.A
>United States
>United States of America
>US
>U.S.

>Hi Anon can you give me all the information about United States?
>Fucking WHERE clause has to include all this shit

>>1726341
Ah you've put forward a good example of ETL for me. I understood the acronym and how data can be put in "wrong" ways but never had a concrete example of how ETL is actually put to use.

Thanks man. Feel free to keep the tips coming if you have the time.

>SQL, Excel, Access, Tableau
What is all this babby-tier bullshit? Where's the Spark, Hadoop, MapReduce, and Elasticsearch? Can any of you fuckers even code?

>>1726364
Why come in here a bitch about coding when most analysts don't do any serious coding if any?

/g/ is the other way.

>>1726341
>>1725891

Walk me through an average day at work.

>>1726378
Depends on the type of analyst. If you're working with any kind of sizeable dataset, Excel and SQL aren't going to cut it.

The big money jobs in data analysis all involve working with very large data sets in real time. You can't do that without some general programming skills.

>>1726417
most data analysts, I would've thought, are engineers or data scientists, so they'd have experience coding

>>1726364
We do have Spark, Hadoop and ElasticSearch, but we have DBAs and Architects who will take care of this shit, I just want my fucking data available so I can NoSQL the HBase with PDI.

>>1726417
Also, real-time is overrated. Just get some Splunk or any other log reading shit live so you can detect major issues, but for business data people are usually happy with minutes/hours late-data.

>>1726355
That was an example of data cleaning, which is ETL so ok, but let's imagine you have 10 websites, e-commerces or whatever. You have some data on Google Analytics, but you also have data on your own database, and you recently bought a database that give you, I don't know, the cities in U.S. who are spending more money on e-commerce. You might want to merge these 3 data sources into 1 thing and use it ot make a report, that is a classic example of ETL. And these 3 data sources have different keys, different information, but at least something that can "join" them. If they don't you'll have to rely yourself on something less reliable (???) like the "name" of a client against another database of clients. But people write names in different ways and people can have the same name, shit like this happens but that's life. You can always use algorithms such as Levenshtein distance to mitigate some of that.

>>1726417
Oh and I agree that you need to understand general programming skills. I was a programmer for more than 5 years, I'm really ok with programming and this gives me a huge advantage desu.

Long story short; data analyst is a meme, too much work, decent salary.
Data scientist is what you want to be.

>>1726964
How does one become a data scientist? I'd also like to know how to become a data analyst?

I currently freelance and I have a strong knowledge of python, rust and some other languages. Not sure what is needed for these fields and what are the usual requirements for them?

Wife's assistant has an MBA and works as a research analyst for her market research dept. She makes about $85K, but she's only been at the job for 5 years. (Ohio) You have to learn the ropes, and an MBA on teaches you the language. Working teaches your to be fluent.

Wife is grooming her to take over the department in 5-10 years. That will double her salary.

Just a new term for a statistician

>>1723711
>tutors
good luck

>>1726341

ETL guy here, that data should be cleaned and standardized before it gets into the db.

>>1724017
Applied mathematics and statistics.

>>1724141
We have a data scientist who does this

They quickly realised DS was a useless meme so the team collapsed, but they kept him so they could tell investors that we're doing mystical visionary shit with data.

Pretty sure he just watches netflix all day, caught him several times now

File: panda-fader-fort-bb10-hot-100-2016-billboard-650.jpg (84KB, 636x421px) Image search: [Google]

84KB, 636x421px

>>1725367
BA here on paper too (sort of marketing automation / python engineer / BI guy)


Pandas is king, fuck excel

>>1724017
Predict elections

>>1725891
>Pentaho

You poor fucker, I hate that thing.

Weirdly, we found that hacky solutions involving a range of services glued together with webhooks and integratiosn soon became way better than any bloatware like Pentaho, never got into it.

>>1726341
See this is why you limit the data users and salespeople can put into your system, they always shit it up

>>1721211
Question
Who's jobs will be automated away first?
>BI fags or fast food workers?

I say BI / data analyst fags in big market regions die first. <10 years. Fast food vending machines slowly transition in, no one notices.

BI and fast food workers continue to exist in poorfag parts of the country (Midwest, flyover land etc) for 20 years but who cares they're poor and gay. Looks like Gummo (1997) all the time there kys lol

>>1727109
Which pretty much means doing some data cleaning even before. There are many fields that are filled by user input and they can go bananas. Also, when working with APIs, Excel, Open Data, etc, you gotta clean or standardise them anyway. Data cleaning is not a meme.

>>1728304
I'm ok with Pentaho Data Integration. I used it and I also used Alteryx before, can't say I miss Alteryx. Maybe I'm ok with Pentaho because I was a dev before so I don't really mind.

>>1728307
I agree. Fuck them. I'm the biggest fan in the world of drop-down menus. You wanna write something yourself? Fuck off, choose one option only. But unfortunately I don't take these decisions... yet.

>>1728390
Wait, you went Dev to BI? How come?

I'm trying to do the exact opposite, have a business degree but programming is the most fun part of work for me.

How do the salaries compare?

SAS master race coming through

>>1728420
Started as developer because of my Bachelor's. Then since I'm not autistic I started seeing that business made more money and became a Business Analyst (functional stuff, dealing with people who made money and communicating their needs to I.T., managing projects). Then I saw people playing with B.I. tools and I thought myself. Did a few nice data investigations using open data, published some, there you go now I had experience and found myself a job.

Data Analyst / BI usually makes more than Dev unless you're a very specific Developer that knows some complex shit. Speaking business I believe B.I. has a lot more potential because you get to talk with CEOs and people who have REAL money. Play your cards right and get rich.

>>1728567
This is literally me now. I'm a business analyst and we're implementing Oracle. Started looking at the BI stuff and it's crazy powerful, gonna do some stuff on the side and impress the directors

>>1721229
This post is a meme.

>>1728328
Please explain how AI, machine learning, and robots are going to automate a BI position.

Please answer. Seriously amused by your fallacy.

>>1728567
It's weird, any time I've seen average DA/BI salaries stacked up against Dev ones, the devs usually make a ton more cash, at least in tech.

BI is interesting, and I like it, but the pay is something that i've been worried about for a while (working at a tech startup for the past two years, work is interesting but pay is shit)

Also for career prospects, what do you do down the line? I know with devs they say if you're still writing code after 30 then you're basically fucked, what happens in BI?

>>1728567
Where did you find the data sets and how did you process those into something that might be useful to a company?

That's if my assumption of you proposing the results of your data processing was handed to a HR type of person, of course.