Day in the Life of a Data Analyst (Work From Home) | *Realistic*
7 Video Views·Oct 22, 2024
A day in the life of a Data Analyst Working from home- this is a video I've been meaning to do for so long.
It is a Saturday.
I have a lot of data analytics work, so I thought I'd do a day
in the life of a data analyst.
Okay, so it's about 8am and I am just going to make some
coffee before I stop my day.
Okay, coffee's done.
I'm just going to go into my office.
So I am starting a new project today, technically a new project.
It's actually an old project, but answering some new questions.
So on my notepad here, I've written the stuff that I need to do.
And in a nutshell, what I need to do is that the client of mine is
asking to answer a few questions.
I have seen the dataset before, and what I need to do is just have a look
at the dataset, analyze it and answer the questions.
What this client does is the client provides a service to the customers
and the service can either be paid by the customer themselves or by the
insurance or maybe a bit of both depending on their insurance cover.
So what the client wants to know is they want to know what
service brings in the most money.
And they also want to know what insurance brings in the most amount
of money and what insurance actually pays the fastest, because obviously
they would want to work with that insurance provider if they bring in
the most money and if they pay quite quickly, there's some other questions
that I need to answer, but for the purpose of this video, I'm not going to cover it.
So the first step here is that I am going to open the data and
just have a look at it on Excel.
So what I generally do is I actually go through every column in the dataset
to make sure that it's fine, make sure that there are no nulls, no empty values.
If there are any columns that I don't understand, then I will just send an
email to the client or contact them just to ask for more information.
So now that I'm done, what I'll do is I'll actually pop the data into Tableau,
which is what is on my screen here.
And I am going to start my data exploration.
So in terms of data exploration, I use Tableau.
Tableau is a popular data visualization tool in data analytics.
And whenever I want to do a quick analysis, develop dashboards, do
my own data exploration, I use Tableau.
You can also use Power BI.
I think Power BI is free, Tableau is not.
You do have to pay for a license, but there is obviously a free trial as well.
And mainly why I like it is because I can just drag and drop the Excel file
in here, or if I am working on a cloud data warehouse, I can just quickly
connect to a cloud data warehouse as well.
So the client doesn't want a dashboard.
They don't want anything automated.
They just want their questions to be answered as it is in this point in time.
If I wanted to do something automated, I would use something called Python,
which I have done an automated report for this client before.
That's if I wanted something automated.
But I don't, so I'm just doing a static report.
So I'll just develop the visualizations on Tableau, probably put them in a
document and send to the client.
So I am just going to do some data exploration now.
So, like I said, the reason why I like Tableau is it is quite drag and drop.
So it's quite user-friendly.
Sometimes I don't want to code because it is a bit too technical.
Whereas Tableau, you just have to drag and drop elements on your page
and, you know, your charts are created.
The only thing in Tableau that you do coding is almost formulas like Excel
formulas, you have calculated fields in Tableau, and that's where you can
create your own custom columns based on mathematical operations, etc.
of other columns.
So now what I'm going to do is play around a bit with this and try
and answer the client's questions.
So my data exploration generally looks like this, and I'm just
showing you example charts.
I can't actually show you the actual data.
And ideally why I do data exploration and the charts are quite simple is
because I myself am trying to look for trends and patterns.
Once that is done, then I can actually choose the charts that I like to show
to the client, and then I'll format the charts, etc.
It is terrible and simple because it's just for my eyes.
And then when I'm ready, I will choose data visualizations to show the client.
Now I am going to take a break, play with my dogs, and then I can build
some data visualizations now that I've done my data exploration.
Okay.
So I have done my data exploration.
And now what I am going to do is start to create the data visualizations
to actually show the client.
So the two I'm going to create are quite simple to answer the
questions that I need to answer.
And then from there, I will create a document that details the answers that I have.
Okay.
So that is done.
I now have two visuals that I will use to answer the questions
that the client has given me.
The first one being what service brings the most amount of
money, which I can show you now.
Right.
So something like this over here, just a simple bar chart with some labels.
So per service, we have the corresponding amounts.
So I think it was also important to actually understand the amount of
invoices they get as well as the average amount per invoice, just so that they
can see what services bring in the most amount of money, but also how popular
that service is, you know, because we can have a service that is billing $10,000.
But you know, you may have only done two clients or two customers, whereas you
can have a service billing $50,000 and you've done a hundred invoices or a
hundred customers, but I feel like this chart sort of gives a good representation
of what services does well so they can have a look at it and pick out the
interesting ones, the next question, which is what insurance brings the
most amount of money also done over here.
So some of the method here, I just have the actual insurance ID.
I then have a total invoice count and then the actual amount and the bar chart
just represents how long each insurance takes to pay in days.
So that very top bar is about 127 days, which is not good.
So ideally what the client will do is have a look at this and see what
insurance bills a lot in terms of amount, but whether or not they take quite
long to pay in terms of the actual bars.
Again, all of this is randomized, so I'm sorry if the results don't make sense.
I just multiplied it by a random sector because this is actual client data.
So now what I'm going to do is actually screenshot these charts and put them
in a document and actually construct an email to send to the client.
What I do in the email is I screenshot the chart and then below it, I just
write my thoughts and my findings.
So that includes the question I'm answering, what I found and some next steps.
So that's what I always do.
I can't actually show you that because it's obviously real data, but my email
for it was about a page and a half of written notes on it, just including
things like useful trends, patterns.
I spoke about the insurance schemes that do really well.
So they pay quickly and bill a lot.
And then I spoke about the insurance schemes that don't do very well.
I spoke about the different services that seem to be quite popular.
So where they are getting good numbers and the average amount per invoice is
high, but I also spoke about services that don't seem to get that great ROI.
So where maybe they're doing big numbers, but the average amount per invoice is
quite low, so those are the things that you would need to consider when you
write your report and now that that's done, all I'm going to do is actually
pre-send and it has now been about six hours in, which I know doesn't feel
like it on a YouTube video, but the rest of the day, I'm probably just going
to chill and maybe watch some TV.
Thank you so much for watching.
I hope this helps and gives you a bit more of a clear idea of what a data
analyst does, especially when working with clients.
Thanks so much.
Please like comment and subscribe.
Bye.

