Intro

hey guys and welcome back to the data analyst portfolio project Series in this episode we're going to take the data
that we have worked on so far in our previous episode and build a interactive marketing analytics dashboard using
powerbi if you haven't watched the previous episodes where we covered the introduction data extraction and
cleaning with SQL and sentiment analysis with python make sure to check those out first you can find all the links and
information in the description below let's start pulling some of the data into powerbi and create our data model
measures and build a dashboard together okay so here you guys can see the dashboard that we are going to make

Dashboard Overview
together and just a quick walk through we have a overview page with some numbers conversion data social media
Data customer reviews and some graphs to support that um a detail page on
conversion details social media details customer review details now of course
before we can make any of this we need a data model we need to get these tables
into porbi we have got to make some calculations so we're going to do that
um but we'll take it step by step I'll I'll take you guys through it now remember as you follow along if you
wanted to use different colors different um different graphs different stuff please please feel free there's uh
there's no set you know rigid stuff here I'm just I'm just taking you guys through how how I've done it and and
feel free to to change it up a little bit um let's move this one out of the
way and then we can continue on this one the first thing we have to do is get some data into parbi so we'll go to get

Loading Data
data and then we will go to SQL Server now we're going to pull some data from the same SQL Server that we have worked
on in the previous episode um so my was
this is the name of my um uh my server and you are going to use
the name of your server so I'll just go portfolio project
marketing keting Analytics so that is the server name that is the database
name for reference sake let me show you guys in in SQL Server management Studio
database name server name the ones that you have been doing your queries on top
of we'll click okay um it might ask you to authenticate here just use the same
authentication type that you have used when you go onto your SQL Server most likely you have used Windows
authentication so so just do that um it's going to show these tables so I'm
just going to check them off and show them to you guys before um what they
look like before and after we we we we use the SQL statements that we have created um not going to f fetch the
geography data because if you remember we we get that by combining geography with customers so we'll take all care
take care of all that here in powerbi I'll check those off and I'm going to load those um they're going to be now
loaded in the porbi as is just as the table was in in SQL Ser management
Studio before we did any SQL uh SQL uh queries on top of those so it's going to
fetch those rows in you'll see we get some tables here on the right side we have in the model view you can see that
it has fetched some tables and auto created some relationships we're going to break those later and then fix this
up a little bit um but we'll go to transform data and let me pull that down
one second there we are um um we'll start off with the customer Journey table um I
have the statement up here I'm just going to go and get the SQL statement for that so I will copy um that one and
you know what just to again show you guys this is the statement I just you
know just copy the statement that I need um from for for that table and you can
see here on the right side there's a source and a navigation applied steps um
so I'm doing this just to show you guys before you can see the table it it it isn't it isn't fixed it we haven't
applied our Transformations yet so you can see the NES are there and there's there's other things that you know that
you'll see in the different tables that we'll fix you can see the nulls there the The Columns that we did in cment
analysis are not here um customers is not combined with geography um in the engagement you know
some of the the naming here isn't correct we haven't filtered out newsletter um view clicks combined isn't
split yet so so there are things we need to do um let's start with the customer Journey table so I'm going to remove
this this navigation step um which uh which means that we only have left the
actual query from the database I'm going to go here and then you can see now here I can now go to Advanced options and
then I can paste the statements that we have for this table so when I click okay
it's going to fetch the table again but now it's going to fetch it with the c equal changes that we had in our query
so you can see now the duration is uh is filled out where it was missing and then
you know the it looks you know the the the the Transformations have been applied so that's great we can change
visit dat up here we can change to date if you want you can see there's uh long
decimals you know if you if you uh set this to a fixed decimal number or uh or
if you want to round it to a whole number up to you I'm just going to leave as this it doesn't really make a big difference but w
customer Journey table so the last thing I'm going to do I'm going to go over here I'm going to rename it to fact
customer Journey great we have our first fact table then I'm going to go back up I
have it up here on my screen I'm going to copy the the statement for the the reviews uh I'll copy that I'll go to
customer views I'll remove the navigation click here advanced and I'll
I'll paste the SQL statement there and then it's going to fetch that table and
it's going to fix some of the the places where the review text had double spaces I'll change the review date to date we
have that great let's go to customers you can see no geography
information um let me just do this SQL statement for the customer joint with
the geography data let's go to source Advanced options same thing just copy
paste that you can see now we have the joined the the join table here no dates
nothing that I need to worry about here that's good um before I forget let's rename this one to fact customer
reviews for the sake of showing you guys if I click customers there I can also do it over here dim
customer and then we have we can move that one down so we have the facts and them together um engagement data whe
my statement for that um let me find it Rel products this
is customer Journey customer views um let me
[Music] see did I open it I don't think I've opened it give me one second guys let me
fix this real quick let me see let me see where is
that statement it should open any moment now there we
are this is our engagement data let me get that um remove the navigation step and
then we'll paste this one great now we have that let's just
make sure the ID content ID campaign product okay text this is not a text so
we'll change that to a whole number um clicks we'll change that to a whole
number engagement date we'll change that to a date great we'll change the name to
fact engagement data perfect and then we have at last but not
least we should have our products table let me see if I can find that
statement there we are I'll do the same thing oh oh that
was the wrong one let me get that step back in you can see that I noticed here I removed the wrong Steps From the wrong
table um dim product and then we'll remove the
navigation and we'll change it to advanc and then copy paste the statement for
the products so now we have the different facts the the different
customers um that is great um before I forget we also o need to uh to get some
data from the the spreadsheet or the CSV file that we have um that we created with our
sentiment analysis let me just make sure that I can uh I can find that file so we'll go
to new source CSV and I'll go one
up and then you can see here's the CSV for that one we'll pull that one in um
um it can it understands it's a a comma um separated
table so we'll get that back in there we can drag it up here fact customer reviews with sentiment here the the
spaces have not been removed so I'm going to click this one and this is just an example of you can also do
Transformations and stuff in powerbi but it's best to do it uh on SQL side or in
a transformation tool depending on you know where you work and how things are but but it is
possible to do a lot of stuff here also um I'm going to just click a double
space and then replace it with one space okay and it's going to add that
step there you can see here on the on the right side it adds steps as you do
Transformations um last thing if we hit click sorry close and apply and now you
can see on the right side the table names are going to change and we're going to get the the results from the

Connecting Tables
changes that we have made into PBI when we want to when we want to work on
things um before we can start with the the the calculations um we also need to
go to the model view um let me just delete this ones um because I want to show you guys
how to do this um from the beginning and so that it makes sense for you guys so we'll delete this
let's just do that it is possible to turn off so it doesn't do this automatically for you
for you it's it's in the options um look it up I'm not going to show you guys
right now but we'll take the the customer table the products table and we'll put the dimensions on
top and we'll have these fact tables at the bottom and then we can start to
create relationships so you can see we have customer ID customer ID we'll grab this one and drop I get down there and
now this is important the the the relationship it's a single way and it's
a one to many for for each one entry in the customer's table there are many
entries in the fact table because one customer can have created different rows
in the fact table so it's a one to many so I'll click save on that you can see it's a one number Direction
many um same with customer ID one too
many and then we have do we have custom ID in this one no we don't here we have
custom ID so we'll drag it there same thing one too many then we have product ID let's find
that take the product ID and we'll put it there one too many on that
one then it's a same here one too many on that one um okay so product ID we have in all
tables which means we can connect this Dimension to all of our facts all of our calculations can be filtered on products
is basically what this means um product ID okay
great so um what we are missing though is let me see if I can find it is we
need a date table um so I'm going to uh
show you guys how to do that there is a script for that which will generate a
table for you so what we'll do is we will do a new
table and you can see now it it up here you can add the name so we'll just call it
calendar and then I'm going to paste the script um I'll pause it a little bit if
you guys want to see what the script looks like sorry um or else I'll I'll also make sure I add it in the in the on
GitHub in one of the files there so you guys can also just copy paste it but just to explain what what it does it it
creates a table um where it takes a range of dates which is defined up here
and then it's going to format all those ranges into different types so we have um dat that's in integer year month um
year month number year month short day of week etc etc etc it's going to create
a bunch of columns which you can use to to connect to these tables and then you
have a a date dimension table also so I'll hit enter on
that see we got a calendar here now we have a date we'll take that date let me
find and we'll connect it to visit date same thing there's one date in the
calendar uh table which can connect be connected to many rows with the same date one to many um date to review
date date to engagement date
and at last but not least we have date to review date so now
we have a data model now you can see that we have X number of tables up here which is filtering down on our facts and
the next step is we're going to go back to the report View and we are going to create some calculations so that we can

Creating Calculations in DAX

start to create visualizations um so the first thing that I like to do I like to create a um
specific measures table for that so I'll just go underscore
calculations and I will load that now notice you can see maybe it's a little
hard to see but you can see the the icon here is a table icon this is a calculated table icon uh there's a small
calculator in front of the table and on this one there's uh there's just one
column um and I have the calculations up here so I'm just going to refer to that on this uh this other screen I'll I'll
type them out for you guys as we go U we can start we can add one called uh let's
just call it uh clicks so I'll go new measure and then
it's going to ask me for that one so I'll go clicks and then sum of clicks which is a column in the fact engagement
data table I'll hit the enter and now notice you can see there's a calculator icon on this one if I right click I
delete column one from this one now there's only calculations in that table
it changes icon to calculator um icon um which you know it becomes a a
calculation table or a measure table as we as we like to call it um so we have
clicks if we drag out a card I can drag clicks onto the field for that and you
can see there and you can see there that I have
done something wrong um there is one column here which I think I haven't changed the typon so we'll go to
transform um let me see if I can find this one this must be um on engagement
where is clicks you can see here it's a uh um I thought I changed these but no
okay um clicks is a whole number views is a whole number another one which would have caused some
issues let's go back let's check now close and apply does it fix
it does it fix it yes it does um so it was a text format it should have been a
number format you can't summarize um you know letters so I had to change it to a
uh to a integer to a whole number um so that works it's great um if we drag out
Year let me see if I can just get that as a slicer instead and then we can
change it [Music] to um let's see slice the settings drop
down now you can see when I hit that the the kpi over here changes which is good
it tells me there's a relationship between the date table and the engagement table so that's good so then
we can add the other um calculations now let me see what do I have I have one for
conversion rate let's do likes first so we're going to do likes equals
sum of and you know if if you're unsure as you go if you have forgotten to to set
the right format you can you can always add them and just you know check as you go it's it's also totally totally fine
um and let's go we can go number of campaigns campaigns I think this is a
distinct count let me just double check so I don't do anything wrong
yes distinct count and then we have campaign
ID we have that one oh typo there we are we can add that we can
just replace this one then we get that um then we have a number of customer
Journeys um copy paste that one it's another distinct
count um we have number of customer
reviews can see that that's fine um rating which is an average you can
see I like to click and right click and go new measure in the table where I want to put it you can also go up here go new
measure and then uh this time it's average of the rating column
to to get the average rating let me see we have
views so let's go views equals sum of
views so that's uh it's also fine great
um then we have one which is a little bit more complex and I will uh I will break it down for you
guys so we'll go new measure it is a conversion rate
we just start like this okay so this one is a little bit different than ones that you have seen
so far um what this one does it combines a calculate function with some variables
so you can see there's a variable total visitors uh where you calculate the number of Journey IDs that have viewed
that that created a view that's how many actually visited and had a view then you have H how many actually purchased how
customer journeys led to a purchase and then I'm taking um I'm just saying that if the total visitors equals zero then
you know just replace it with uh with zero just to show a zero in the in wherever I use this measure otherwise
divide total purchases by total visitors and when I'm dividing here I'm referring
to these variables that have been defined up there so that one is a little more
complex than the other ones this is Dax it's it can get quite complicated but I've decided to choose some of the
simpler ones now just so that everyone can follow along and we can we can get through this uh together so now we have

Designing Overview Page

calculations we have a data model and we are ready to start to build some of the parts of the dashboard so what I like to
do is I like to start and just put a nice text box on top just to give it a
uh the page a name I'm just going to name it overview I like to get this one you know
a little bit bigger maybe both it um you know this is the part where you can decide a little bit for yourself how you
want this to look but but this is how I like to do it um I'll select the the
background I will let me see if I can go over here to the format pane canvas uh
sorry canvas background and then select gray and put it to 50 so that I get you
know some some some uh separate separation between elements on the can
and the the actual canvas itself it's just a personal uh personal preference
then I'm going to add a slicer so we can filter different parts and I want to filter on product name so I will add
that here then I will go up here you can click here add data I know that you can
also you can do it here if you want to you know I'm just used to going to the search and and searching for the stuff
that I want so it's it's some of a habit but of course you can also just go here
and then add data and then search for it here it's it's totally up to you guys what what you guys prefer
um ah I need to actually add it let me make sure it's selected there add data
and and there we are we have a drop down so I can select different parts and then uh and then uh filter the the the
whatever I put on the canvas basically um so we had three different parts we had convers
social media and customer reviews so I'll add a text box I'll write
conversion and then I will put that let's see I will put that one I'll put that
one give it some space maybe something like that and then put in the middle
pull down a little bit I like to make some separation so I'll put it right next to it and shift and right uh you
know one one to the right and I I get that we can even make it smaller to be honest so we have the conversion
section oh and then we can have social media let
me put that there social media
media and we will move that down
there maybe one more great and then we had custom reviews and that's what's
good is that when you have the data model you have the measures then things starts to go a little bit quicker um so
custom reviews great and we can add a card for
the conversion rate so I'll put that same size one to the right I'll find my
measure I'll go to calculations conversion rate can see that it
is a rate like that if you want to you can click it change it to percentage if
you want to show it as a percentage also totally fine um and then I'm going to
I'm going to select it crl C contrl + V put it under here and then I will add
three boxes for views
and me see make sure I have enough room so so so you can see it's a little
bit big now and I've done this on purpose because I want to show you guys something um we'll go to format and
we'll go to call out value we'll put it to let's say we put it to 30 little bit smaller and we can put the label to 10
and to not have to redo this over and over and over we can now go to format painter we can just paint onto it and I
can see it's still a little bit big but it's fine because I'm going to change
it um and then we'll have one down here
which is rating tell you what I'll just do it like
this I think it'll be fine I might change it later let's see um so now
everyone says conversion rate let's find you know we have views click this one I'll add it there
and we'll fix that also in a moment clicks great
likes perfect and then rating oh make sure we activate this one
and rating great so these ones um still a little bit too big let's go 25 better
let's go 20 oh try 20 that's fine display units
none okay format painter we get the same effect great the decimals on these ones
are uh it's hard to read the numbers as it we'll go to views here we'll add here
we'll add a uh a th000 separator on these ones easier to read so we'll go
there same on the likes now I can already see that these
are these are values for for all years so we need to fix that
um so we'll add a uh we'll add a slicer um let me find it here and we'll
take the the year into that we'll change it to
a to a tile great let's turn the slicer header
off put it like this can make it nice and small and then we'll take
that and we we'll add month name short on to that one you
can see it's sorted a little bit weirdly but let's start by clicking 2024 that's
good um you know it's it's it's working so far so good we can go to month name
short here sort by column and then you find month number it's going to sort it
um you know ask the numbers you know January number one February 2 March Etc
3 four five etc etc um okay good so we have some kpis we can click around you know things
are starting to to happen we can click over here you can see there is
interaction um now we have calculations we're starting to get a layout we can start to add some
visualizations um so we'll take conversion rate I'm going to show you some different ways to to create
visualizations now I'll go contrl c contrl v so we double it uh make some
room for it and then I'll click over here um to a line chart um you know you can
always uh go here also and you can you can change it here to a line chart if you want to but we need to add a month
name so I'll search for month name short because we need to have it across some sort of
axis uh let me make sure I have this selected on the x-axis then we get this
part great contrl + C control V go back in um I'll change this to a uh a
uh let's try a clustered column chart now then we can find the product name
and we remove that and we add that instead okay it's it's a little bit tight maybe we
can make it a little bit smaller let's give it a little more room
um the x-axis let's give it some more space for the the labels it's still a
little bit tight but but it is what it is it's uh it's fine it's fine we'll
we'll uh we'll just have to go with that for now um down here we have views clicks
and uh and likes so we'll go contrl C actually this one I'll just show you
guys from scratch you can go to build over here you can right click go add
Visual and then just you you you'll get the uh a slot for the visual right away
you click line chart and then you can go on the x-axis I know we want month name
short um on the Y AIS we want some calculations so we can go likes sorry
clicks likes and Views that will give us those over time and then um let me
see uh and then this is
fine uh it's a little bit hard to see the months we're going to clean this up a little bit give me a little bit of
time here let me see well make sure I do this the right way if we put this one over here okay that gives us those we
can go to the Legend um can turn off the title we don't need that um this is fine um let's
just do some small changes now let's make the title a little bit smaller it's a little bit big um if you want to you can turn on
the data labels you can get a little bit noisy but okay let's let's keep it like that for now on this one I'll turn the
labels off going to sort it as sending instead I think that that's
better [Music] um again we need to turn off the title I
think that's a little bit more room at least okay I think that's fine like I
said there's no you know there's no 100% right answer you guys can also play around a little bit you know get some
get get a bit of feel for how it is to work in this and and and pay attention as you as you go to how you you shape it
um we're going to create uh we're going to have views then it's going to go to clicks and then how many likes we get
you can see here we need to do something with the labels because you can't read that so we
go uh data value and percentage of first and then let me see if I can
figure um if I can figure this out category labels
those are fine let's see conversion rates they also look
uh look fine there we are that's the values
display in none okay that gives us the actual values also so that's good um
okay so we have used clicks likes we can see some of the the conversion that goes down on uh on that also um last but no
least we will go to Let's
see we will go to this section which is rating so I'm just going to grab the
same graph remove conversion rate and then add ratings average to see that over time and uh same as this one I'm
going to apply some changes let's see if we can get some of the
some space for this one it's just a little bit too tight let's just remove
this so we give it a little bit more space give me one
moment okay we'll pull that over there we'll move you over
here there we are and that should be enough okay and then at last I had
ratings um part chart and then I believe I have
product name is what I think I have over here which is also you know it's also
quite uh quite tight uh for the names and that's what
uh that's what happens when uh when uh we we get more and more
elements on here um of course we could go in you know you can play around with some of the sizes on the x-axis you can
make it a littleit smaller same on the Y AIS um on the values just to see if you
can squeeze out a little bit of uh space there you can give the labels more um
more uh space Also that's also you know doable but it is going to be a little
bit tight when um when when you look at a whole year but if you go in and select one month it it becomes a little bit
easier um let's also make sure we sort this one as sending then where you want to you can
turn on you can turn on labels if if you would like
that so already we have quite a bit going on here you know you can go and you can select your product name you can
see how that filters down you can select one month or hold control and select 3
months you can see how that filters down there's there's quite a bit of insights and and flexibility already
here just by creating this first overview page um and what I like to do is I like

Designing Conversion Details Page

to now go from an overview page into more detail pages so we can drill into
certain parts so the next one that we had is uh is is conversion
details so so you can see here it's it's just you know the conversion rate and
then just spread out on on some other factors and I just restructure it uh a
little bit just to to to just focus in on one thing so this is the the finished
version um that that I have and and we can do the same thing so what I like to do I like to duplicate the page so I
don't have to recreate everything I'll give it a new name conversion details
give it a new name up here conversion details and then I can start
to remove things I don't need on this one so we can remove this one can remove that one we can also select
and remove several um so what I would do here is I would um give this one a
little bit more space and we just change it like this
just's make sure that we have it the other way and then we can turn on the data labels now you have you know a lot
more space um so we can we can see them now nice and clear let's make them a little bit small
smaller they still quite big um I think that's fine um and the last thing I had
is I had the let me see let me see I had one I can remove this
one get in the conversion right here I
will let me see I got to make some room for that other graph like
this so want to make sure we have enough room for that that's good I'll move you over here and then I had conversion
rate per product so I just grab the the the measure I I change it to a matrix so
you can see the value still has conversion rate so you know I I'm trying to just work as efficiently as as
possible here product name that gives us that we have it sorted by the conversation rate up there that's good
month name short um we can add that to columns and we'll see that spread out you can see the
table becomes a little bit wide now because we have a lot of decimals so we can go to conversion rate um do we need
two decimals maybe maybe not I'm going to take it down one notch um you can even take it down a a full and just
keep the the full percentages now you can see we have uh more space um depends
on the requirement and and what you what you think makes sense uh maybe one uh
decimal is good to have um and then we want to take a look at how many customer
Journeys have been taken through different uh uh took different actions
as they convert down so we'll add this this funnel graph just put them next to
each other and then let's see if we have enough space um we have number of customer
Journeys on the values action you know what action do they take in their customer Journey
you can see that it it it breaks down uh you know I like this type of graph when
we're looking at the conversion I think it shows it in a smart and easy
way um that was not what I was hoping for it it fixed a little bit and it also
messed up some other parts um let me see data labels values
15 not even 10 much better um um and then data value
and percentage of first and one decimal place is good
enough actually let's keep that one
at there we are that's better great so
now we have our first detail page um it focus a little more on the conversion
rates so you can go in and you can see more details on different elements if
you are interested in that um the next part is going to be a social media

Designing Social Media Details Page

Details page so then we will um I'm actually going to duplicate the first page so that I don't have to redo too
much stuff social media details we will go in we'll remove some
of this stuff I will keep let me see I'll keep
this [Music] one that one and
this one I'm actually going to remove this one you can keep it hey keep it if you want um I'm going to move this one
um over here and then we'll add one
more which is similar but I'm going to change it to
a cluster column chart and then instead of having uh clicks and links I think I'm going to
go actually it will keep views and then we will add content type
so now we can see clicks and likes and view over time and we can also see views within the different types of contents
over time um you know another thing that I think is
good then um I wanted to create one table which shows C similar to this one
so I'm going to go back and grab this table contrl c contrl v get that in here um and I'm
going to focus a little bit on the views to understand which products um are getting a lot of views so let me just
move this one up there let's just adjust a little bit like that and we'll remove
conversion rate and for the sake of the the
exercise I'll I'll go over here um add data on values calculations and we'll go
views it will show number of views per product over time um we can sort it by
the total to see the product that gets the highest views and then if we find content type then we can we can add that
as a row here so you can go into different products and look at within the
different content sorry within the different products how did it perform in different content type types um if you
want to you can even change the views to be a uh a percentage either of the row
that it's on or in the column total um so you know if you have 100% uh down
here in January in baseball glove you know how much do that make up up to 100%
or or other um ways that you want to do it it's just for me to to show you guys
some of the different possibilities here but we'll we'll keep it like this right
click condition formatting background color um lower values we can give that
red higher values we can make that you know blue add a middle color which
is actually let's find a green color for
for higher something like that so now you can see that the higher
numbers they get a green value and the lower they get a a uh a red so if you go
in and you select May then you can try to or I mean it's easy if you have one
month but if you have two or three months then you can you can see how it it changes a little bit to show you know
which which uh which are doing very well and which are doing not so good and even if you expand then you can see within
different product names which uh which um content types have done good within
those products so it's just me once again just trying to show you guys some different uh possibilities please play
around please try you know shape this however you would like
to um the last page is customer review

Designing Customer Reviews Details Page

details so we'll go in here and we will go customer
review and before I forget I don't think yep we have to social media
details go here customer review details perfect um and the kpi there
was the the average rating so now I'm just going to pull them in one by one
um just to show you guys how that is done yes it takes a little bit more time but it's it's nothing too nothing too
crazy we'll make this one a little bit smaller um and we can see that we
have let me think I'm trying to look up here at the same time while I'm working
down here um number of customers per rating and and this is where the sment
part is going to come in a little bit and I'm going to show you guys why um you know how how that creates value here
but we need to measure number of uh customers let me just make sure I have
customer reviews actually yes there on the Y AIS and then we're
going to have the rating score on the x-axis um let me make sure I have this
right this one it's because I built a measure on on this rating not this one and we have
two um there we are and then we need to make sure that it
treats the X AIS since it's a number as a
categorical that looks better let's make it a little bit
larger great sort as sending and then we can turn on
the detail labels so this is basically telling us how many people gave
different uh different scores make sure you sorted by the rating by the way yes
so you know 26 gave a 1 27 gave a 57 33 gave an 88 Etc
um let's just grabb some of the formatting from this one so we going have to redo
that and then which means I need to turn on data Labs again but that's done quickly ah sorry about that there we are
um so so that tells one side of the story now that we have the sentiment uh also we can go in and we
can take one of the columns that we created in the sentiment analysis so you can see now
we have let's just check this this is the sentiment I think it's a sentiment
category yes so we'll do the same thing sort this
one and we'll sort it ascending so you can see that even
if so even if 88 gave a uh A3 you can see there's some mixed sentiment within
those 88 um the positive one ones were you know mostly of course positive the
82 um the the the 82 negative you know there's degrees of negativity so the
sentiment gives a little more insight into that into that part so I think
that's a very nice uh way of kind of like I said enriching the data with a
different perspective Without Really um changing you know what we are
calculating or or how we're doing it we're just adding some more um um some more value to the data so we can get a
different aspect and understand it a little bit better um so so I think this shows a little bit of the value of of of
doing that um then we can take the the sentiments and we can see how they've
developed over time so we'll we'll have that we'll change it to a line chart I'm
going to for month name short add that to the
x-axis and then we'll take the sentiment category and add that to the The Legend and you can see how the sentiments have
changed over time um and then we had one table which
gives some details on the different customer reviews which were given so
we'll give that some space let me just fix that so the fields that I added there was
uh that was review date so we can go review date and then we have the
customer ID the review text the sentiment category that it falls within
and the review I believe the
reviews uh rating of course and we can turn off the sum we don't need it to
summarize that so don't summarize and then give it a little bit
more space that should be good do we have enough room let's let's keep it let's keep it
like that instead we can try to see if we can go to the presets we can
go uh play around let's put a minimal I
think maybe if we yeah if we make the values a little
bit smaller we'll have enough room uh what this means you can go in you can select a rating and you can see
all the different comments that are under that rating uh now what is maybe more interesting is to go in and look at
the the negative or the mixed um mixed negative or mixed positive to maybe
contact those and see if we can do something about that um you can see it is also increasing so that's something
that we want to take note note of um just some just some different ways to to to Showcase this uh this data um let's
also give this this label a little bit more room so so we can see it um fully
so now you can read it fully that's that's good and the last one that I want to show you guys is a little bit more
advanced um um but we'll add a a scatter chart here we'll put okay so we'll put
the average rating that was given we'll put that on the x-axis and then we will
see you know how many customer reviews do do we get per different types of uh
ratings so we'll put that on the the y- axis now of course this just become the intersection of uh of both of this uh
but if you add if you add the customer ID on this one then it it spreads out a
little bit now because there's just a lot of uh dots there right now if we
then add let me show you guys if we add the the customer
uh let me find it the sentiment category as the legend you can now see that it
starts to it starts to break into different almost like different clusters
let me just grab this format painter and and add that formatting um and then you
can start to see you know how how many uh customer IDs do we have per different
uh uh different let's call it's almost like average or or rating groups just to
see where where do the different reviews um actually fall and another thing that
I did is that I took the let me make sure I do this correctly uh based on
number of customer reviews uh I want the size to to be big based on
that also so that it's easier for you to understand so so basically it's the
bigger the marker the the more the more uh customer
reviews uh came in on that that point so it's a little bit more of a
advanced visualization if you can call it that um let me see if I can give this
this Legend some more so we can actually see it now we can see
it let's do it also here now we can we can read them
fully um okay but but what we're seeing
is that you have different ratings one 2 three four five and then upwards is how many uh you know how many gave you kno
how many on this rating did you have in customer reviews so of course you want as many as possible here on the right
side and towards the the upper side these are the ones that are in average these ones that are down left are the
ones that you want to you want to try and catch and do something about so so to go in and see okay if I click this
dot you know which reviews for that if I click this one which reviews for that
click this one which reviews for that um that's some that's some parts that that
you can you can you know do to kind of see it in a a little bit of a different
um visual way another thing that we we can look at is that you know you can also use these slicers and and go in and
then you can add the the sentiment category if you want
to uh category on to onto this one so that
if you want to focus on only one sentiment category okay who were the positive ones who were who were neutral
who were who were negative mixed positive mixed negative maybe those are the ones that you think Ling turn and we
can also add that as a slicer and and go into those those details um so that is
the that is how I created the customer reviews details so now it's up to you
guys you know try to you know look at the video play around create the measures create the data model pull it
in and then you know take a look at what I've done take some inspiration from that and then then try to do this for uh
for yourself um I will upload the finished powerbi desktop file on GitHub
um the blank one I think you guys can make yourself I will add the the script
for the date table I'll add that on GitHub so you can see that also um of
course even though if I upload these sources I highly highly recommend that you guys use this to practice test adapt
it a little bit the way you want to do it uh so that you get some of that practice and you get some of that first
you know hands on EXP experience on how to build a a interactive dashboard in in

Next Episode
porbi that's it for our episode on building a marketing analytics dashboard with powerbi in the next episode we're
going to focus on presenting some of the insights using PowerPoint where you'll learn more about how to effectively
communicate your findings to stakeholders make sure to like subscribe and hit the notification Bell so you don't miss any of
series if you have any questions or thoughts then leave a comment below and I will get back to you guys there other
than that thank you guys for watching and I'll see you guys in the next episode