Kiva Project Overview
| Previous | Current |
|---|
Logistics has always been a huge industry, and even more so with rapid technological advancements since the 21st century.
Given XXX being one of the largest companies in the logistics and transportation industry, it faces stiff competition from its competitors from other major companies such as Fedex and UPS, as well as local and regional companies who are more price-competitive in their region of expertise. Contracts with corporate companies in particular, take into account several factors when choosing their shipping partner, such as efficiency of delivery in terms of time taken, and this in turn includes other factors such as the ports which XXX will use for the delivery and the mode of transportation, such as air, ocean and roads.
Being students in the Analytics track, this project provides us the opportunity to utilise the skills learnt from the theory taught in school and practice them from an actual dataset out in the industry. We hope to understand and gain insights into how XXX maintains its competitiveness, with factors such as service, competitive pricing and brand power affecting the bid price of its contracts, which in turn affect its revenue and profit.
The main objective of the project is to help XXX better understand the various reasons behind the final bidding result for their contracts. We aim to explore and determine trends behind the bidding prices at different times of the year, and identify factors behind the differences in pricing, by analyzing into factors such as seasonality, different bidding habits of companies in different industries, differences between customers’ expected prices and actual bid prices, differences in expected and actual shipping times and so on.
Some of the salient questions we have include:
- What determines XXX's winner ability in the market? Apart from competitive pricing, to what extent do factors like service reliability (in terms of speed of delivery), expertise (varying in different countries and territories) affect the overall revenue?
- Many different sectors of customers are involved, such as Technology, Life Sciences, Retail and more. How do the different sectors decide on their buying criteria? What is the difference in price sensitivity across the different sectors?
- To what extent external, uncontrollable factors such as seasonality, natural disasters or big events affect the sales?
- Does the location which the company resides in allow them to exhibit stronger brand influence in the region? (Differences between XXX Singapore, XXX China, XXX USA and so on)
Our main objectives are:
- Identify possible clusters based on sectors of customers or requirement of customers in terms of speed of delivery or special services.
- Identify possible seasonal patterns of customer target price and bidding results using time series analysis.
- Identify possible relationships between customer’s sector, speed of delivery and other factors which affect the bidding results.
- Identify patterns of the difference between customer target price and bidding results, and patterns of the difference between customer required transit time and XXX provided transit time.
- Provide meaningful visualizations which explain and account for the bidding habits of different customer profile groups
- Develop a predictive model for XXX to be better prepared for future contract bids and optimize their pricing strategy for greater profit.
Exploratory Analysis
An exploratory analysis will be conducted to help us understand the shipping patterns and bidding results of different customers.
- Identify the major inbound and outbound region, country and city for each customer
- Understand which transportation mode customers prefer
- Identify patterns of the difference between customer target price and bidding result
- Identify patterns of the difference between customer requested transit time and XXX provided transit time
- Study the relationships between different factors and the bidding results.
Customer Analysis
We plan to categorize customers into different groups based on the relevant variables identified in exploratory analysis (referring to exploratory analysis point 5). Each customer will be assigned to a cluster based on this analysis and we attempt to have a better understanding of the buying behaviour of the customers.
Time Series Data Analysis
Because each excel file is only valid for a certain time period, time series data analysis could be conducted to discover the underlying seasonal trend in customer target price and bidding result.
