ANLY482 AY2017-18 T2 Group15 Data Visualization Platforms

From Analytics Practicum
Revision as of 16:14, 15 April 2018 by Amanda.lim.2014 (talk | contribs)
Jump to navigation Jump to search

HOME

PROJECT OVERVIEW

DATA ANALYSIS

DOCUMENTATION

REFLECTION

ABOUT US

BACK TO MAIN ANLY82


 

DATA VISUALIZATION PLATFORMS USED


Using the data which had been reformatted into Flat Table format, dashboards were created using two data visualization platforms - Qlik Sense and Tableau. These platforms were chosen as they were suitable for the client's needs.
All charts presented in this wikipage have been created using randomised substitute data to protect our client's confidentiality.


THE CURRENT DASHBOARD


Client's current Brand Dashboard in Excel
Figure 3: Brand Dashboard in Excel


Figure 3 is a snapshot of the client’s current dashboard for one of its main product brands, which was generated using their original data structure.
This dashboard is split into two graph segments, namely the input graphs and the output graphs. The two input graphs on the left are of pre-sales data and include the distribution of brands across the various distribution channels, and the prices of their products across those channels. The four output graphs on the right are of post-sales data and include the equity and imageries of the brand as provided by customers, the awareness, penetration and usage of the brand among customers, and the sales performance plotted against the sales volume.


This method of data visualization may not be ideal for several reasons.
  1. As the data was neither in Relational nor Flat Table format, it is not possible to automatically load the files into visualization platforms requires the data to be input manually, which takes unnecessary time and effort.
  2. The resulting visualizations are highly static and lack filter options, making it difficult for users to slice and dice, to view a specific segment of data – a function which is readily available in other visualization platforms.
  3. Monthly addition of new data into the dashboard will be a time-consuming process due to the manual input method required.


To address the three issues above, our team has first converted the client data into the Flat Table format which will be easily loaded onto alternative visualization platforms, then replicated our client’s dashboard as seen in Figure 3 onto other platforms more suited to our client’s needs. To improve the flexibility and adaptability of the final visualization, we will prioritize features which support the seamless addition and integration of new data into existing visualizations.
The amended design of the dashboard in Figure 4 includes a filter to allow customization in the data displayed. It will also be dynamic and interactive, allowing data segments such as date, visualization colours, and graph selections to be changed with a few simple clicks.


Dashboard Design Sketch
Figure 4: Dashboard Design Sketch


The team created this dashboard on two platforms which it feels will better suit its client’s needs: Qlik Sense and Tableau.


Data Visualization Platforms Used


QLIK


Our first recommendation for a visualization tool is Qlik Sense. Qlik ranks as one of the top data visualization platforms and a choice solution for many large Multinational Corporations (MNCs), offering improved aesthetics and more user-friendly design processes than primitive visualization software.
TABLEAU


Tableau is a popular beginner-friendly visualization software which is similar to Qlik but holds additional advantages for our client.


QLIK DASHBOARD


Brand's Dashboard in Qlik Sense
Figure 5: Brand's Dashboard in Qlik Sense


The Qlik dashboard shown above showcases an improvement in design and aesthetics as compared to the Excel platform. Similarly, the incorporation of filters is an upgrade to the original dashboard. Furthermore, for Qlik Sense, through data associations in its data loading process, filter selections can be applied across all datasets for synchronised changes in all the charts.


TABLEAU DASHBOARD


Brand's Dashboard in Tableau
Figure 6: Brand's Dashboard in Tableau


Unlike in Qlik, the Tableau sheets can be combined into a single dashboard and resized to fit the user’s screen, which was a request specifically made by the client. Dashboard creation and customization is also simpler due to Tableau’s drag-and-drop-based format, which is more intuitive than Qlik.
One simple but very effective example of Tableau’s unique functions is the expansion of the date field. Given that the data is in the correct format, the user can easily choose the specific time frame to display data. For example, Yearly data can be expanded to Quarterly, or Monthly and so on, with a single click, as shown below. This is a function absent in Excel and Qlik.


ADDITIONAL CLIENT REQUESTS


In addition to the graphs in our client’s current dashboard, they have also requested for specific adaptations to the Yearly Awareness, Penetration and Usage (APU) graph to be created, shown below.
Brand's Dashboard in Tableau
Figure 7: Client’s Original APU Graph (Yearly)
Firstly, the client requested for the above yearly graph to be done in quarterly format. This was done in Tableau, shown in Fig. 10 below.
Brand's Dashboard in Tableau
Figure 8: Revised APU Graph in Tableau (Quarterly)
Secondly, the client requested for two new ratios to be plotted: the Trial Rate and the Conversion Rate, explained in the table below
Caption
Trial rate Conversion rate
The rate at which customers aware of Brand A go on to try their products. Given by (Total Usage) / (Total Awareness). The rate at which customers who try Brand A go on to become regular consumers. Given by (Regular Users) / (Total Usage).
Brand's Dashboard in Tableau
Figure 9: Brand’s Trial and Conversion Rates in Tableau
Brand's Dashboard in Qlik
Figure 10: Brand’s Trial and Conversion Rates in Qlik