Data Analysts Job Description

mtb_sends

Hi All,

Just wondering if any data analysts out there could describe what they do on a day to day basis (especially analysts in the public sector/civil service). Just trying to gauge what the work would be like to do. Its a career I would be interested in pursuing and also feel it would suit my character profile.

ELM327

Data analyst is a very broad description which is used by a lot of companies to describe varying roles.

What are you looking to get into, what is your background etc. I was a data analyst a few years ago before moving to data science and then analytics.

mtb_sends

@ELM327 Yes I've noticed the job descriptions for the role seem quite broad. Background is in software development. I suppose the data wrangling aspect and generation of infographics are the parts that peak my interest the most.

What are some of the core differences you noticed between the three roles you mentioned above. What were the pros and cons of each from your point of view?

Flinty997

Most of the ones I deal with aren't that technical from an IT point of view. They have tools like Tableau, and PowerBI, and then R etc, and use those to analyze their data and produces charts from those. Many of those tools have image functionality to produce the graphics. Infographics are more something a graphic design produces with those graphics under the direction or the data people. Data analysis is a specific skillset, you can't really wing it. While the IT development technical skillset is very useful in data. Its not the same.

mtb_sends

@Flinty997 What is the character profile you think is best suited to a data analyst role if most are not that technical?

Flinty997

Attention to detail and methodical. Don't make assumptions.

A lot of IT people are not great at seeing the bigger picture, or thinking outside the box. They often don't analyze a problem before tying to solve it with some technical solution. They won't look for a non technical solution for example.

https://www.spssanalyticspartner.com/blog-what-makes-a-good-analyst-8-pointers/

https://webbiquity.com/web-analytics/6-traits-of-highly-effective-data-analysts/

Flinty997

That's just my opinion. As EML says its a wide area. What suits in one job might not suit in another depending on the situation and organizational culture.

Will_I_Amnt

As a Data Scientist who is a former data analyst, I never really liked the Dashboard building side of it with Tableau, Qlikview and Power BI. Much of my role as a DS uses my skills as a DA also - They say 80% of a DS role is data sourcing and prep - if anything, it's more than that. Having at least an understanding of Data Science practices is of growing importance in the role of a DA. Familiarity with at least one cloud platform such as AWS. Azure and Google Cloud likewise as well as data platforms that sit on top of them such as Snowflake. Having a background in software development should be useful as you should be already very familiar with things like JSON structures, SQL and traditional relational database designs.

mtb_sends

@Will_I_Amnt & @Flinty997 Thanks for the detailed responses, definitely a few points to think about there. Having read plenty of data analyst job descriptions to date, a lot of them do seem to be more geared towards technical candidates in my opinion. I'm open to correction though.

Flinty997

Most of the places I've worked, Data people don't have free access to the data. They get a sanitized very limited access to a data export specifically for them. But everywhere works differently.

mtb_sends

@Flinty997 Out of curiosity was your experience of data people in the public sector or private sector?

Flinty997

https://www.boards.ie/discussion/comment/119304092#Comment_119304092

The joys of recruitment. They always over spec job requirements.

That said knowing more is mostly always better.

mtb_sends

https://www.boards.ie/discussion/comment/119304165#Comment_119304165

I always wonder if organisations that produce these ridiculously extensive job descriptions actually narrow their candidate pool by turning good people off the role. I've lost count of the number of times I've decided against applying for a job based on the requirements of the job seeming so far fetched and unrealistic.

Flinty997

Just part of the process. Apply anyway. Test the truth of the spec.

Will_I_Amnt

https://www.boards.ie/discussion/comment/119304092#Comment_119304092

Well pulling data out of a database is a technical skill. Even throwing it into Excel and building pivots or marcos on it is technical. But making sense of it also requires an understanding of the business process that creates and then alters the data - which varies in complexity. Depending on the size of the company and its legacy systems, you can touch on an awful lot of things.

mtb_sends

@Will_I_Amnt In your role of a DS, have you come across many data scientists coming from the one year data science masters programs most colleges and uni's are offering?

Will_I_Amnt

https://www.boards.ie/discussion/comment/119304358#Comment_119304358

I have only worked for one company as a DS and most other DS's in the company are not in Ireland. We hired a 4 of them in Ireland 3 years ago - all of them are gone because there is a fairly big jump in salary for someone who spends a couple of years out of college as a DS and then moves on. But choose your college wisely. Try go for a course that has matured over a few years rather than some new fad that a college just started this year or last year. One with good practical, hands on content rather than tones of mathematical theory. In interviews, I've seen the poorest candidates coming from the same colleges all the time.

mtb_sends

@Will_I_Amnt I've already completed a degree in it to be honest, a couple of years ago. I was not impressed with the standard of the content, it was first/second year undergrad level in a level 9 course in my opinion. I noticed some companies/organisations specifically asking people who have completed these level 9 data science programs not to apply which I can completely understand.

Will_I_Amnt

https://www.boards.ie/discussion/comment/119304462#Comment_119304462

I did a 2-year part-time Masters in DS at TUD (it was Blanchardstown IT when I started and it was TUD when I finished). It was decent enough. Depending on what supervisor you get for your thesis, I'd recommend it.

callaway92

Roles I offer often are titled 'Data Analyst' and like other have noted above, it's usually on Power BI that we work with

Our 'data' is data that comes from coding sports (mainly Gaelic Games). That data exports in a CSV format and we look for ways to display and analyse it nicely then

It is a very broad description of a job title though, and often it's an Entry Level title

StupidLikeAFox

At a minimum you will be required to gather, collate and summarise data. However, if you want to be a good data analyst you will want to provide insight, or more importantly actionable insight

Lots of people confuse data analysis with statistics. Using @callaway92's example, plenty of people in GAA think they are doing data analysis when really it's statistics (which are useless on their own). They might say "you had 20 possessions and 10 turnovers", when you want to be saying "you have had 10 turnovers (data), 70% happened after short kickouts and 10+ handpasses (insight) therefore we need to kick longer passes to the full forward line (actionable insight)".

It really depends what the employer is looking for though, data analyst could be compiling reports or else answering broader, open ended questions with data

callaway92

https://www.boards.ie/discussion/comment/119304676#Comment_119304676

Exactly - Lot of people in GAA circles call themselves data analysts but generally they are Data Coders. They aren’t analysing the data. Quite frustrating

Flinty997

https://www.boards.ie/discussion/comment/119304315#Comment_119304315

Might be technical from an end user (Admin Worker) point of view. But from an IT perspective its basic stuff. But as you say it might be vastly more technical depending on the organization. But like you say. It's how long is a piece of string.

Business or domain knowledge is important as you say. But the principals are the same. But we find you have to agree metrics even in the same organization. As different business units define things differently. Even in the same organization.

Flinty997

https://www.boards.ie/discussion/comment/119304160#Comment_119304160

Data analysis as you think about it today didn't exist with that title when I started in IT.

Most of my recent experience (with data people) is with public sector bodies. I'm usually providing the data or explaining it. The data is used for past, current and future analysis. Reporting, predicting trends, informing policy and strategy.

But I don't do any of that "analysis" myself. I'm on the IT side of this equation. I have dabbled in it from time to time, just for kicks.