Hello, I’m back with Dashboard week – Day 2’s blog.

The data for Day 2 is Melbourne City Council Data guys and girls! We were given a task to solve the business question of where to open a new café in Melbourne City?

The main data set for this challenge is Café, restaurant, bistro seats — CoM Open Data Portal (opendatasoft.com) and it was advised that supplement dataset could be added in to deepen the analysis.  The most interesting part is that we have to visualize our analysis using PowerBi tada!

THE DATA

The data from City of Melbourne Open Data Team is already nice and clean. There is information about the data and dataset schema which explains all the fields. I could also preview the data in table format directly on the website, which is nice. These data are for open use, so it is ready to be downloaded, by accessing the Export tab.

THE STORY

This is one of the challenging parts for me in this project. I could not find anything interesting from the original given data set, so I decided to enrich it with additional data. There are a lot of ready to use data sets from Melbourne Open Data, which is good but also quite overwhelming as there are too many directions.

A good tip I found is to take a step back from the data and allow myself some time doing research on the business question, which had helped me to have a clearer direction of which data set I would like to add on.

I added in data about employment for Melbourne city area to find out which areas have more workers and low number of cafes. Jobs per space use for blocks — CoM Open Data Portal (opendatasoft.com). I joined the two data sets using the Block ID from both sets.

THE POWERBI VIZ

It’s all about regular customers! I decided to take an angle of finding a location for the next café which will have lots of potential regular customers.

This type of customer will usually live or work within a short walk (5-10 minutes) from their regular café.” (The 7 Things to Look for in a Cafe Site (sevenmiles.com.au))

I managed to build a few charts to tell this story. I must admit it is harder for me to analyse in PowerBI. There are so many things I thought I wish I could do this in Tableau :P.

Below is the image of my finished dashboard:

After building this dashboard I have reviewed/ learnt the followings:

  • Changing data type for geographic data
  • Editing interaction for filtering/ highlighting
  • Creating new measure
  • Creating dual axis chart and label formatting
  • Background and canvas formatting
  • And a lot more fail attempts to do things in PowerBI 😛

THE ANSWER

So, it’s a wrap for my dashboard week – day 2, thanks for reading!

See you in the next blog.

Thao

The Data School
Author: The Data School