Product Sales Report
All code is in Producr Sales Analysis notebook
Background
- Pens and Printer launched a new product line of office stationery 6 weeks ago.
- As launching is expensive, we need to make sure we are using the best techniques to sell for each new product.
Methods:
- Email: Customers in this group received an email when the product line was launched, and a further email three weeks later. This required very little work for the team.
- Call: Customers in this group were called by a member of the sales team. On average members of the team were on the phone for around thirty minutes per customer.
- Email and call: Customers in this group were first sent the product information email, then called a week later by the sales team to talk about their needs and how this new product may support their work. The email required little work from the team, the call was around ten minutes per customer.
Questions
- How many customers were there for each approach?
- What does the spread of the revenue look like overall? And for each method?
- Was there any difference in revenue over time for each of the methods?
- Based on the data, which method would you recommend we continue to use? Some of these methods take more time from the team so they may not be the best for us to use if the results are similar.
- We don’t really know if there are other differences between the customers in each group, so anything you can tell us would be really helpful to give some context to what went well.
Data Validation
The dataset contains 15000 rows and 8 columns before cleaning and validataion.
I have validated all the columns against the criteria in the dataset table:
-
week: 6 unique values without missing values, from 1 to 6, same as product launching timeline. No cleaning is needed. -
sales_method: additionally to 3 original categories had 2 misspelled categories, which were mapped according to the original ones. -
customer_id: all values are unique with the same lenght without missing values. No cleaning is needed. -
nb_sold: numeric values without missing values, from 7 to 16. No cleaning is needed. -
years_as_customer: numeric values without missing values, however 2 values - 47 and 63 - don't make sense, as counting from our foundation year 1984 they are in the future. Imputed with the number of years from 1984 to today - 41. -
nb_site_visits: numeric values without missing values, between 12 and 41. No cleaning is needed. -
state: 50 possible values without missing values. No cleaning is needed. -
revenue: numeric values with 1074 missing values completely at random. As there are multiple different products, the revenue will vary depending on which products were sold. Unfortunately, we don't know product for each row. So if we want to make non-biased conclusions, it's better not to impute them now. Missing values were dropped.
After the data validation, the dataset contains 13926 rows and 8 columns without missing values.
How the Sales Differ Between Different Sales Methods?
Half of all customers were communicated via Email and three times bigger than Combined method, which has the least number of customers. Call method is two times bigger than Combined method. Considering that Call method takes more time than Combined (30 minutes call vs 10 minutes call on average), there are a lot of customers for Combined technique to save more time if needed.
But how the sales differ between different sales approacher? Revenue is one the most important measures of sales method efficiency, so let's see which approach works best. We see a wide spread of revenue distribution having several peaks, which could mean different groups of data.
Adding sales methods on revenue distribution, indeed shows us that each method has different revenue range. From the boxplot below, Combined method has the highest values of revenue, while average revenue for Call method is twice less than for Email.
Besides the biggest average revenue, Combined method also have higher average number of sold products, based on the bar chart below. Email and Call methods have about the same average number of products sold. Also, considering that the Call method also requires more effort and has less average revenue than the Email, it doesn't make sense to continue using Call approach any further.
Is The Combined Approach Worth The Effort?
Email method has the biggest total revenue, which is not a surprise, taking into account the number of customers it has. On the other side, Combined method is definitely more efficient than Call method.
For the same numbers of products sold, Combined method has higher revenue comparing to other approaches. Thus, Combined method seems more efficient to sell higher-priced products. While Email method, which takes a very little effort, can be good for promoting mass-market products.
Are there any changes over time? From the product launch, average revenue is steadily increasing for each sales method. So each one becomes more efficient over time.
At the same time, total revenue for Email method is steadily decreasing. For the last 2 weeks, Combined method has more total revenue, which makes this approach the most perspective.
Business Metrics
Since our goal is to have the most effective sales strategy for the new product line, I recommend to trace several metrics.
- Percentage of Total Revenue per Sales Method
This metric shows the proportion of total revenue generated by each sales method. Thus we can quickly identify which sales approach is most effective for the new product line. Comparing trends over week can show if effectiveness grows or declines.
Week 5 shows that Combined method is almost 50% of all generated revenue. Next week it's predominant even more, covering 68% of all revenue.
| week | Call | Email + Call | |
|---|---|---|---|
| 5 | 21.0 | 29.2 | 49.8 |
| 6 | 17.3 | 14.5 | 68.1 |
- Total Revenue per Sales Method
This metric shows the total revenue generated by each sales method and adding context for us. By comparing absolute revenue across methods, we can not only quickly identify which sales approach is most effective for the new product line, but also our sales performance overall. Comparing trends over week can show if sales grow or decline.
For the last 2 weeks, Combined method generated most revenue, however sales are declining.
| week | Call | Email + Call | |
|---|---|---|---|
| 5 | 53518.11 | 74373.45 | 126809.72 |
| 6 | 28252.92 | 23706.75 | 111152.07 |
- Weekly Growth Rate per Sales Method
This last metric complements the previous one and tracks how the revenue by each sales method is changing from week to week. It shows the momentum of sales and whether sales tactics are having a positive impact.
Email method revenues has been declining for the past 2 weeks, while Combined method sales increased significantly in week 5, and decreased slightly in week 6.
| week | Call | Email + Call | |
|---|---|---|---|
| 5 | 7.9 | -31.0 | 62.0 |
| 6 | -47.2 | -68.1 | -12.3 |
I believe that together these 3 metrics can show the overall picture very quickly without having to dig deep. Combining Email and Call approaches in one method looks promising, so we should keep track these metrics next 4 weeks before the sales strategy meeting to ensure that Combined method is indeed the most effective.
Recomendation
For the following weeks, I would recommend we can focus on the following steps:
- Spending less time on Call only method, and more on combined Email + Call method
- Using key metrics to monitor the most effective sales approach
- Data Collection for in-depth analysis
- Improve data quality - Recover missing revenue values
- New related data - Adding product types to compare sales strategies between products:
- Combined method with high revenue, but less customers might focus on premium products.
- Email method with many customers, but lower revenue might target mass-market.