Market Analysis of iFood in Excel
Created by: Bryan Eckard
Utilization of DoorDash is very high in this household. It is very convenient for a family of 5 who has various commitments at, often, conflicting times. While placing an order the other day, I began wondering if other countries had services like this, and if they spend as much as we do on food delivery.
So, after some online research, I found a dataset about a company in Brazil called iFood. Thanks to this data, I was able to gain some key insights into how iFood's sales and demographics work.
A few of these insights include:
Customers spent a total of R$1.24 million (Brazilian Reals) (~ $242,837 US) over a 3-year period.
On average, customers spent R$187.59 per year.
67.7% of the spend variance can be explained by income.
Customers in the 51-65 age bracket spent the most.
The Data
The dataset I used was actually given as a data analyst use case scenario for an interview process. The data is for the years 2014-2016.
Here is the link to the data: Marketing Analytics | Kaggle.
A single row seems to represent one customer, their purchasing habits with the sevice, when they joined, and various other demographic information. The main columns from the dataset I focused on were Income, MntTotal (Total Amount), Age, Customer_Days, and DateJoined. The dataset had been slightly modified for educational purposes. The data dictionary included did not state what currency was used. Therefore, I have assumed it to be the Brazilian Real.
The Analysis
I chose to use Excel for this analysis because the dataset is small, and it has all the tools I needed. To start the analysis, I noticed the dataset was relatively clean, but did not have a unique identifier for each row. So, I created a customer ID column and assigned one to each.
To better understand how much of their income each customer was spending, I created a new column and calculated what percentage of their income was spent using the service. These were all relatively low aside from one outlier at 70.66% of their income. This may be due to a typo in either the Income or MntTotal columns. Then, I created some summary statistics.
Next, I created a scatterplot to compare income and total spent. This showed a positive relationship with a relatively close grouping.
I inserted a trendline and calculated the R-squared score which was 0.6774. This shows that approximately 67.7% of the variance in spending can be explained by the income. As can be seen, there are a couple of extreme outliers on each side which need to be investigated further.
Finally, I created an age group column using the provided age of each customer. I then inserted a pivot table to perform some analysis based on age.
As can be seen in the above table, the age group spending the most was the 51-65 bracket. This table was subdivided into individual ages as well. Customers aged 50 years old spent the most at R$48,337.
I believe the sums are more useful because the averages are skewed due to there not being an equal number of customers in each age bracket.
Final Thoughts
To recap, a total of R$1.24 million was spent over the three years of the dataset. Customers spent approximately R$187.59 on average per year. About 67.7% of the variance in spending can be explained by income. People in the 51-65 age bracket spent the most.
With this information, I would suggest that iFood spend more on marketing to wealthier areas and customers aged 51-65. These may not be mutually exclusive as most people make more as they get older and more experienced.
Thanks for reading! If you have any questions, feel free to submit a message below, reach out at my email (ch13f_48@zoho.com), or connect with me on LinkedIn Bryan Eckard.
Comments