ANLY482 AY2017-18 T1 Group03/Midterm Progress
| Home | About Us | Project Overview | Midterm Progress | Final Progress | Project Management | Documentation |
For our initial analysis, we mainly did exploratory analysis to have a better understanding of the data at hand. This included the basic analysis of distribution of each column in the data files to observe the similar trends and the extreme outliers.
After cleaning the data and getting a better understanding of the data files, we decided to run exploratory analysis on products, sales channels, therapy areas, sales rep and customer interaction types (Phone Web & F2F), revenue and sales quantity.
We also did the Pareto Chart analysis to find the top 20 products which are garnering the top 80% of the sales to see whether it validated the company's top 2 products.
Moving forward, we plan on using sensitivity analysis and further analytical tools to help us dig deeper into the data so that we can find actionable insights for our client.
Visualize Data and derive Insights about Sales Sensitivity to Customer Interactions:
To fulfill the client’s request to determine if factors like interaction duration affects sales profile (quantity and revenue), we intend to carry out sensitivity analyses as characterized by each factor. Such factors include interaction time, the presence of Phone/Web vs F2F type of interaction, regional distribution (postal districts), etc. We will be visualizing the data as deemed fit by our scope of work. After observing trends through visualizations, we intend to report derived insights and recommendations through a final project report and presentation.
Create Dashboard to report Insights to Manager:
For easy periodic reviews of the company performance by the client, we would like to explore the creation of a dashboard on the Qlicksense platform. This will perform the task of generating (a certain set of) insights from every additional duration’s sales performances entered into the system. The steps required for this are:
(a) Preliminary Wireframing / Brainstorming of Dashboard elements:
Based on the current progress of work, we now have a better understanding into the workings of the company. We also know which fields are common across data sets (allowing for better cross-file aggregation of data). Lastly, we have also been in communication with the client, who has expressed a general desire for insight into sensitivity on sales.
With these, we aim to create a basic trial “view” of what a potential dashboard could look like, with elements which we think would aid the client in making operational decisions, or deriving certain insights. This should be done before our next meeting with the client.
We also plan to do an internal review of whether some of the Dashboard elements / functions are achievable given the skill level of our team. This knowledge will help us in revising the Dashboard along the way with the client (replacing technically tedious requirements by rearranging certain information, or communicating to the client how each substituted data insight presented on the Dashboard links to the overall update she hopes to piece together)
(b) Initial review of dashboard element:
At the first (after the mid-term) meeting with our client, we aim to get the client’s direction on some of the elements we suggest in the wireframe. During this meeting we will explain various reasons (why we feel we would be unable to provide a certain functionality) and rationale (why we are suggesting it be done this way) behind the preliminary wireframe. As we begin work on the platform, this review would continue on.
(c) Finalising Dashboard:
We strive to complete the Dashboard by the first week of Nov (Week 12). This is to give ample time for revision and troubleshooting.
(d) Troubleshooting:
Throughout the creation process, we will be troubleshooting each element on our end (both technical as well as data related) to ensure robustness of (i) our system and (ii) our data analysis framework. This final period of troubleshooting is catered as a deadline for our team to present the final product to the client, and if time permits, for the client to explore and experiment (based on her requirements for her scope of work) on the Dashboard. This gives us time to respond and work any changes into the final product.
