OList Marketing and Retail Analytics: Capstone Project A Brazilian E- commerce Company

OList Marketing and Retail Analytics: Capstone Project A Brazilian E- commerce Company.

Objectives Background Visualizations Insights Recommendations Appendix – Data Sources Data Methodology Data Assumptions.

Objectives. To identify top products that contribute to the revenue and top product category using Pareto Analysis. Improve our understanding with the use of market basket analysis to analyze the purchase behavior of customers. Understand what items are most likely to be purchased individually or in combination with some other products..

Background. OList is an e-commerce company that has faced some losses recently and they want to manage their inventory so as to reduce any unnecessary costs. Now to be able to meet the demands of the customers, the company would need to store tons and tons of products in warehouses. Since storing these products adds to the costs that the company incurs, it is necessary for the organization to plan their inventory well..

Top 20 Ordered Products by Quantity. Sample Footer Text.

Top 20 Ordered Products by Revenue. 5b50f2e2 6cddS38 d6160fb7 dlc4270 99a4788e 3dd2a17 250855 5f504b3a 53b36df6 a 07 d eOd64dcf d28S360f 7a1078r nc7f353 f819f0c8 588531'8 422879e 16c4e87b 5a848e4 a62e25e 63.885 0 47.215 43.026 41.083 38.907 37.734 37.683 37.609 31.787 31.624 30.468 29.997 29.024 28292 26.577 25.034 24.229 24,051.

Percent Running Totals. Percentage Running Totals by Revenue and Orders Product ld bb50fze236e5.. 6cdd5384349.. d6160fb7873t„ dlc427060a0f.. 99a4788cb248.. 3dd2a17168ec.. 2*38557cf79., 5t504b3a1c75.. 53b36df67ebb.. aca2eb7dOOea.. eOd64dcfaa3b.. d28S360f29ac.. 7a107g16372.. fic7f353075c.. fg19fOcg4a64f.. sggs31fgea7.. 4228/9e10f46.. 16c4eg7b9ga9.. sagage4ab52E. a62e25eoge05__ 2tA609f8948b„ fd0065af7f09.. a5215a7a9f46,_ bc4cd4da98dd„ Revenue sg.ggs 54.720 48 47.215 42.026 41.083 38907 37.734 37,683 37.609 31, 787 31.624 30,468 29.997 29.024 28.292 26.577 25.034 24.229 24.051 22.717 22.000 21.740 21 soo % Of Total Running 047% 087% 123% rgo% 220% 248% 2.75% 3.04% 3S5% 3.78% 401% 503% 521% 571% 587% 603% Total Quantity % Of Total Running Ordered Quantity Ordered 0.19% 0.25% ogg% 0.70% 1.54% 2.30% 2S0% 262% 277% 2.92% 297% 2.99% 3.35% 3.37% 3.56% 3.74% 4.00% 4.01% 402%.

Revenue Pareto. 247.526 garden_to 1107,saa 193.4" oet_snoo 160.326 comouters lss.864 nome_aool m 146.854 138.021 1317S3 nome_appJ 12S.643 124.169 luqoaoe_• 117.352 u 115772 111.544 industw_c 9.759 a 276 rood 4.ggs 1.572 928 ot Total Sum.

Quantity Pareto. nn'tn_tm '2791 s 12,036 11.721 42 computeß food 9 3 % of Sum of ot.

Product Category Ordered >5 Times. Category Ordered More Than Five Times E097 11.01S 4gs 40k 45K of orders 50K.

Market Basket Analysis. Market Basket Analysis product Category N a me (Order! Prod I) bed _ bath _ table furniture_decor _ CCe SSOr_ _ health_beauty watches_qi fts sports _ leisure housewares g a rd _ t OOI S auto telephony stationery coo _s tuff pet _ shop baby o ff ice_ fu r n Product Category Name (Order!Prod2) 8 21 294 156 104 21 21 71 21.

Insights. The category Toys constitute 20% of the products which generates 80% of the revenue. It can be seen that even if the price of the certain products is high, it is still bought by the customer more often. Apart from Toys, the products from the categories of bed_bath_table, furniture_decor, computers accessories and health beauty are the most frequently ordered. The above categories with Toys or/and with each other are most frequent in customers’ basket. It is observed that despite of the high price, some products are frequently purchased by the customers..

Recommendations. The company should focus on the categories which generate more than 80% of the revenue by always keeping them in stock. The company should target customers who are more likely to buy toys to boost sales as the category toys is the most ordered category. Offer promo-codes or discounts on the frequently ordered category associations to encourage cross selling among the products. The company can reduce some of the sub categories which have very low sales..

Appendix - Data Sources. Here is a snapshot of our data dictionary: Order details such as order id, order status, order purchased timestamp, etc. Order Items detail such as order item id, seller id, price, shipping charges, etc. Customer details such as customer is, customer city, customer state, etc. Payment details such as payment type, payment value, etc. Product details such as product id, product category name, product dimensions, etc. The following data sources were used: OList retail dataset containing order-related information. The data consisted for the year 2016 to 2018..

Appendix - Data Methodology. A thorough analysis of the OList Retail Dataset was conducted. The process included: The dataset was cleaned and transformed using the python libraries of Pandas and Numpy in the Jupyter Notebook. The missing values for the various columns were replaced with the best values. The redundant and duplicate records were discarded and only first occurrence is kept. Exploratory data analysis was done using the python libraries of Matplotlib and Seaborn in the Jupyter Notebook. A new dataset consisting of order id and product category name was created for Market Basket Analysis. Various visualizations and Market Basket Analysis was conducted in Tableau..

Appendix - Data Assumptions. Only the cases having order status as ‘delivered’ are considered. We assumed that the data provided was achieving the desired revenue. We assumed that the company does not want to expand to new warehouses. The company’s strategies are decided considering there is constant growth in sales..

Thank You.