Performance Analysis of PetMind's Sales Strategy
Introduction
PetMind, a pet retailer, sells items from 6 categories that they have broadly classified as either luxury or everyday products. They have been wanting to increase the sales of food and other everyday products by increasing their repeat purchases. After testing this strategy over the last year, they now want to analyze the effect of repeat purchases on their overall sales. They need information on their catergorical repeat purchases, overall sales distibution, and the relationship between sales and repeat purchases.
Data Prepartion for Analysis
PetMind have shared their stores’ sales records for analysis. The dataframe contains 1500 rows and 8 columns. The values in each variable have undergone the following data validation and transformation operations to align with the provided data descriptions before we began our analysis:
What can we observe from the volumes of repeatedly purchased products across different categories?
The data revealed that more than half of the 1,500 sales transactions were for repeat purchases. This would indicate that PetMind’s approach seems to have positively impacted sales. Proportionally speaking, equipment contributed to a quarter of these transactions while medicines, housing, food, and toys followed with 16% – 17% each. Accessories had the smallest number of repeat purchases, suggesting that consumers were generally least inclined to purchase them more than once.
Note: To capture all the data and make our categorical aggregations of repeat purchases a little more meaningful, I apportioned the tiny sum of records that we tagged as "Unknown" to each of the 6 defined categories based on their respective weightage.
What can we understand from the overall distribution of PetMind’s sales?
PetMind's sales distribution is right skewed, which basically means that most of the transactions were concentrated in the small to medium revenue range, with fewer transactions as we look to the higher value end of the graph. The least profitable sale was a purchase of accessories for medium sized fish which sold for USD 286.94, while its most profitable one was a sale of toys for large birds, which sold for USD 2255.96. This is a large range between the minimum and maximum sales values.
Looking at the dispersion of all sales over the period, it would be quite reasonable to try to identify and study transactions of distinct products that brought in sales between USD 350 to USD 1,650 as they were about 95% of all recorded sales transactions.
How did sales of repeat purchases perform against non-repeat purchases? Is there anything that can be inferred from these findings?
While 60% of transactions stemming from repeat purchases would imply that they outperformed non-repeat purchases, revenues of sales brought in from non-repeat purchases generally had higher values. This is given by 2 factors that are visible in the graph:
- An average transaction value that is 5% higher than that of repeat purchases
- A wider and higher IQR (range of middle 50%) and overall spread of transactions
✅ Key Takeaways...
This analysis revealed that pet equipment sells better than the other categories of repeatedly purchased products. While looking at overall sales, transactions for each product had a wide range with an average sale value of USD 996.60. This average didn’t change much when observing the repeat vs non-repeat sales so it would be safe to use the general average as a ballpark measure. Repeat purchases did occur more often than non-repeat purchases, so PetMind's approach does seem to be fruitful. However, further analysis of performances by other variables such as pet type, product size, and product ratings, as well as a mix of different variables is strongly recommended for better determination of specific products or product lines that PetMind could concentrate their efforts on. They can also experiment with price changes to increase revenues of repeatedly purchased items.