代写Big W - QBUS6600 Project Outline帮做Python语言程序

日期：2024-03-17 10:41

Big W - QBUS6600 Project Outline

Background

BIG W (https://www.bigw.com.au/) is part of the Woolworths Group.   Operating nearly 180 stores nationwide and employing over 18,000 people, it is a discount department store focused on selling general merchandise and everyday goods.

The  discount  department  store  market  is  highly  competitive;  BIG  W’s  major  traditional competitors are Kmart and Target.  BIG W also competes with the supermarket business unit of the Woolworths Group and the Coles supermarket chain.

BIG W’s core customers as a discount department store tend to be price sensitive and will often cross-shop with competitors in an effort to find the best offer for the products they are buying.

In recent years, BIG W’s customers have been struggling due to the prevailing economic conditions, with the cost of living increasing and the effects of inflation. They will typically have less disposable income and less purchasing power than in previous years and, therefore, will be highly discerning about where to spend their money. This has resulted in a tougher trading environment for BIG W.  BIG W’s stated purpose is to “make a real difference for families” - to this end, Woolworths Group aims to find a way to sell more products to BIG W customers whilst maintaining their gross profit.  This goal will allow the business to deliver more value to their customers without  damaging their  profitability, which  is  of particular concern  in the aforementioned difficult trading conditions in Australia.

Mechanics of Pricing, Promotions and Availability

Pricing

BIG W has a national pricing model, which means that the selling price for an article is the same in every store.

•         The following limits are applied to the frequency and amount of price changes to satisfy legal requirements and limit the burden on stores for changing product tickets:

o    The maximum number of total price changes (total number of products with a price change applied), excluding promotional price changes for a single week, is 20.

o    Price changes are applied on a Tuesday and should be assumed to take effect for all transactions onwards including those on the Tuesday, until such time as another price change is applied.  Any price change must remain in effect for a minimum of six weeks.

•         The provided competitor pricing data is not exhaustive.  Changes in price should take this into account.

Promotions

BIG W runs “promotions” for specifically identified articles (products) for limited durations in partnership with the vendors supplying the product.  Note the following on promotions:

•         BIG W  runs promotions (discounted pricing) on a regular basis.   These promotions generally have a 2-week duration and run from Tuesday to Monday.

•         BIG W receives funding from vendors to support promotions.  This funding is allocated at a brand level.  Predictions of vendor funding should be consistent with the supplied historical data.  See ‘scanback’ in the supplied data dictionary for related information.

•         There must be at least two weeks without a promotion or price change before a new promotion is activated.

Availability

•         Not every article is sold in every store.   You are provided with a daily count of how many stores are selling each article (otherwise known as having the article “ranged”) and a count of the stores both selling the article and with the article in stock.

•         Stock availability for each article will vary by store.  For the sake of simplicity, you don’t need  to  consider  constraints  in  supply  that  may  result  from  price  changes  and promotions.   However, you may want to consider historical supply constraints whilst developing your strategy.

Problem Description

You  have  been  provided  with  a  dataset  from  BIG  W,  relating  to  the  sales  and  profit performance for a selected set of Categories, along with data collected on competitor prices for  products  directly  competing  with  these  BIG  W  products.    The  data  provided  spans approximately two years and is detailed in the “Data Description” section below.

Please carefully review the “Mechanics of Pricing, Promotions and Availability” section above to  understand  how  pricing  and  promotions  operate  and  what  limitations  exist  on  these mechanisms.

In addition to this dataset, you are encouraged to explore external datasets to enrich your analysis and feature engineering.

In this project, you will:

●   As a business analyst, you will do a preliminary Exploratory Data Analysis (EDA) of the dataset. You are expected to find or reveal all possible properties, characteristics, patterns, and statistics hidden in the datasets.   The results from your EDA may be used for the final goal of identifying the metrics and relationships that are useful for predicting daily sales volume (units sold) and, subsequently, gross profit.

●    Synthesise your  potential  insights from the  EDA and construct a model which can predict the sales units and gross profit response to pricing and promotion changes.

You will need to build a model with whatever machine learning approaches you feel appropriate. You should evaluate your model/s on a range of metrics; however, the RMSE (defined below) will be used to evaluate the performance of your final model on the test data. You should follow an industry-recognised approach to Data Science problems (e.g. CRISP-DM) and include a justification for your selected model.  You will be required to show the methods you used to prioritise your potential insights and defend the models and results with supporting evidence. You will also be required to submit your sales predictions on the test data.

Important:

1)   Please use the pre-split training and test set that has already been provided.   Your evaluation metrics on the test set are important.

2)   Please  consider which  variables are not available at the time of fare look-up, and exclude those as predictor variables (because in real life, your model won’t have them available when making predictions!).  You can read more about data leakage here:

https://www.kaggle.com/code/alexisbcook/data-leakage

●    Based on your analysis, design a potential project for the Woolworths team to execute, to take advantage of the sales, pricing, promotion and competitor data available to increase the volume of units sold to their customers without decreasing the overall gross profit.

This project could include (but is not limited to) a pricing strategy or optimisation of promotional activity.   Each  project must be supported by an estimated improvement in overall volume (units sold) and the corresponding predicted change in total gross profit with supporting data and assumptions. Woolworths Group plans to run the project as a test over thirteen weeks, from the week commencing Monday 05/02/2024 through to the week commencing Monday 29/04/2024 inclusive, so the group should focus on recommendations for this time period.

Please limit the number of recommended projects to 1-2.  Also, note that it is ideal for groups to recommend the deployment of their model.   However, groups can also  leverage model insights for recommendations, as long as the recommendations are closely linked to the insights and not overly general in nature (e.g. general app redevelopment or event).

Data Description

The data provided spans approximately two years and is detailed as below:

● BIG W Sales and Prices: contains ~1.85 million rows altogether.  Provides daily sales and profit data for each product sold in a selection of BIG W categories from February 2022  through  to  January  2024.  The  data   has  been  pre-split  into  a  training (FinSalesPriceData_train.csv) and testing set (FinSalesPriceData_test.csv).

● Competitor Price Data: contains  ~77.7k  rows.     Provides  weekly  level  data  on competitor prices mapped to BIG W products from the week commencing Monday 14/11/2022 through the week commencing Monday 29/01/2024 (inclusive).

The provided data in both files has been limited to four product categories:

●    Household Cleaning

●    Baby Consumables

●    Personal Hygiene

● Skin & Sun Care