DEV Community: Gabriel Wainaina

Building an Interactive Excel Dashboard for E-commerce Product Analysis: A Case Study of Jumia Products.

Gabriel Wainaina — Tue, 08 Sep 2026 15:42:06 +0000

Introduction

In this project, I set out to analyze a dataset of products listed on Jumia, one of Africa's leading e-commerce platforms, with the goal of understanding how pricing strategies, discounts, and customer reviews influence overall product performance. Using Microsoft Excel, I cleaned and prepared the raw dataset, engineered new features to categorize products, ran correlation and trend analyses to uncover relationships in the data, and ultimately built an interactive dashboard to present the findings in a clear, business friendly format.

The objective was not just to explore the numbers, but to answer practical business questions: Do bigger discounts actually drive more customer engagement? Does price influence how a product is rated? Which products are genuinely performing well, and which ones might need a different pricing or marketing approach?

This article breaks down the full process behind how I turned raw data into a finished report from initial data cleaning to the creation of an interactive dashboard.

The following is the uncleaned data(original dataset)

Data Set Overview.

The dataset contains information about products listed on Jumia with the following columns:
•Product: Name of the product.
•Current Price: The current selling price of the product (in KSh).
•Old Price: The original price before discount (in KSh).
•Discount: The percentage discount offered on the product.
•Review: The number of customer reviews received by the product.
•Rating:The average customer rating of the product (out of 5).

Data Cleaning and Preparation

On a separate sheet after copy pasting the original data set I had to ensure there was no any present duplicate in the dataset.

Now that I had removed any duplicates that was present, next process was changing the cell format for each column from General to the appropriate data type i.e. Text for product names, Currency for prices, Percentage for Discount and Number for ratings and reviews.

Next, I used Find and Replace on;

Both the Current Price and Old Price, replacing "KSh" and replacing it with blank.
On rating category to correct it from Ratingd to Rating.
Replacing the out of 5with blank in the review category.
Replacing blanks with unknown

On both the Current Price and Old Price columns, there were highlighted cells (as seen in row 40 shown in the image below) containing a price range instead of a single value. To correct this, I used the =AVERAGE() formula to calculate the average amount on each cell.

The Review numbers were in negative values and reviews can't be negative hence so I used the formula =ABS() to remove the negatives.
I also used the =TRIM() and =CLEAN() functions on the Product column to remove extra spaces and non-printable characters that could interfere with sorting, filtering, or matching later in the analysis.

Below is the final Cleaned Dataset.

Data Enrichment.

I Created additional columns which would improve my analysis.

Discount Amount.
Formula used-
=(Old Price − Current Price).
Rating Category.
Formula used- =IF(OR(F2="Unknown",G2="Unknown"),"Unknown",IF(G2<3,"Poor",IF(G2<=4,"Average","Excellent")))
Discount Category.
Formula used-
=IF(E2<20%,"Low Discount",IF(E2<=40%,"Medium Discount", "High Discount"))
Price Category.
I used QUARTILE.INC() to create named breakpoints that split my Current Price data based on its actual distribution, rather than relying on arbitrary cutoffs.

I created two named ranges:

Price_Q1 → =QUARTILE.INC(Cleaned!$B:$B,1) which is the 25th percentile, meaning 25% of products are priced at or below this value.
Price_Q3 → =QUARTILE.INC(Cleaned!$B:$B,3) which isthe 75th percentile, meaning 75% of products are priced at or below this value.

Using these two breakpoints, I then classified products into Low Price (bottom 25%), Medium Price (the middle 50%, between Q1 and Q3), and High Price (top 25%) using the formula:
=IF(B2<=Price_Q1,"Low Price",IF(B2<=Price_Q3,"Medium Price","High Price"))

The final Data set

Data Analysis

I used the formula =AVERAGE() to find;

Average current price of products= KES1,187.
Average old price of products= KES1,811.
Average discount percentage= 37%
Average product rating= 3.9

I used the formula =COUNTA() to find the total number of products which is 112.
To find the sum of total number of reviews I used the function =SUM()which is 723.
To find the most and least expensive products I used =MAX() and =MIN() which are KES 3,750 and KES 38 respectively.

Trend Analysis

I wanted to find out whether;

Higher discount percentage result in more customer reviews.
Using=CORREL(F2:F113,H2:H113) the correlation was r =-0.1368.
This means that there is a weak negative relationship between discount percentage and number of reviews. Hence as discount increases, reviews very slightly tend to decrease.
Highly rated products receive more customer reviews.
Using =CORREL(I2:I113,H2:H113) the correlation between Rating and Review count was found to be 0.0572, indicating virtually no relationship. This suggests that review volume is not driven by product quality, but likely by other factors.
Expensive products rated higher than cheaper product.
Using =CORREL(B2:B113,I2:I113) the correlation was 0.1100 indicating a very weak positive relationship. This indicates that more expensive products tend to be rated very slightly higher, but the effect is negligible. Price does not appear to be a reliable predictor of product rating.

while also;
• Identifying the top 5 products with the highest ratings.
Using the filter option on rating, I was able to filter the rating column to see the top 5 products as shown below;

• Identifying the top 5 products with the lowest ratings.
Using the filter option, I was able to find the top 5 products with Lowest Ratings.

Product Performance Analysis.

I was able to analyze the following using the filter option in the respective categories,

- Top 10 products with the highest discounts.

- Top 10 products with the highest number of reviews.

- Top 10 highest-rated products.

- Products with high discounts but low ratings.

- Products with strong customer engagement.

To identify products with strong customer engagement, I created an Engagement Category column that classifies each product into one of three tiers i.e. Weak, Average, or Strong which is based on a combination of review count and rating.
Rather than picking arbitrary thresholds, I used quartiles function
QUARTILE.INC()to set data driven cutoffs based on the actual distribution of review counts.
I created two ranges:
Review_Q1 — the 25th percentile of review counts
Review_Q3 — the 75th percentile of review counts
Then I used the formula: =IF(OR(H2="",H2="Unknown",I2="",I2="Unknown"),"", IF(AND(H2>=Review_Q3,I2>=4),"Strong Engagement", IF(H2<=Review_Q1,"Weak Engagement","Average Engagement")))

Then I filtered the Engagement column to retrieve Strong engagments as shown below;

Seller Performance Analysis.

Using the data i was able to answer the following by simply using filter.

- Which products appear to have strong customer demand?

- Which products may require better pricing or marketing strategies?

-Are there products receiving many reviews but having average ratings?

Using the Quartile Formula i was able to comeup with a review category and I used the formula =IF(OR(H2="",H2="Unknown"),"", IF(H2<=Review_Q1,"Low Reviews", IF(H2<=Review_Q3,"Medium Reviews","High Reviews"))) to highlight the three tiers respectively hence allowing me to see the products being asked.

Hence deducing yes.

-Are there products with high discounts but low customer engagement?
Yes.

Dashboard Design.

Business Insights and Recommendations

At the end of this project, the following insights were drawn from the analysis:

1. Are higher discounts leading to higher customer engagement?
Not significantly. The correlation between Discount and Review count was -0.1368, indicating a weak negative relationship.This suggests that discounting alone is not an effective driver of customer engagement on this platform.

2. Do highly rated products have higher or lower prices?
There is only a very weak positive relationship, with a correlation of 0.1100 between Price and Rating. Ratings appear to be driven more by product quality or customer experience than by price point alone.

3. Which products are performing best based on reviews and ratings?

4. Which products may need improved pricing strategies?
Products which have weak engagements.

5. What recommendations would you give to Jumia sellers?

Don't rely on discounting alone to drive engagement. Since discount percentage and review count showed almost no meaningful relationship, sellers should invest in other areas rather than assuming bigger discounts will attract more attention.
Price is not a strong signal of quality perception. Moderately priced products can achieve ratings just as high as expensive ones, so sellers should not assume that premium pricing alone builds customer trust.
Focus on products with Weak Engagement. Sellers should investigate why these products are not attracting reviews or interest as this represents the biggest opportunity for improvement.
Encourage more customer reviews overall. Since review counts were generally low across most products, sellers could benefit from actively prompting satisfied customers to leave feedback, as review volume does not appear to be strongly tied to any single factor in this dataset.

GitHub repository link: https://github.com/wainainagabriel63-dev/Jumia-Product-Performance-Dashboard

My First GitHub Project: From a Local Folder to GitHub Using Git and SSH.

Gabriel Wainaina — Sun, 23 Aug 2026 06:37:37 +0000

Introduction

The following is a step-by-step process of how i was able create my first GitHub project from a local folder and upload it to Github using Git and SSH.

Creating a local folder using git.

Using Gitbash, I wanted to see the list of files and directories on my pc. I used the command function ls which enabled me to see them.
Since I wanted to create a folder in Desktop under OneDrive, I used firstly cd OneDrive which allowed me to access OneDrive then used cd Desktop allowing me to access Desktop. Basically the command function cd (change of directory) allows one to access a directory that is required while Using cd .. allows one to go to the previous directory.
Using mkdir Kenya-Hospital-Health-Records-Project I was able to create a folder under the name Kenya-Hospital-Health-Records-Project. We can use hyphens, underscores and quotes in names that contain more than one word because inclusion of spaces will create separate folders instead of the required one folder.
Using cd Kenya-Hospital-Health-Records-Project to access our folder, I created another folder inside it called data using mkdir data.

Creating a README file

I created Readme file using touch README.md where .md denotes mean Markdown. A Readme file would be able to explain the context of our project on Github and to do this we use the command function echo to print text.
I used echo "# KENYA HEALTH RECORDS ANALYSIS" > README.md to indicate title of the README file and using # to indicate it is as the main title.
Using echo "## Project Overview" >> README.md I was able to mark it as a sub heading. using > instead of >> would have simply replaced our Tittle an used the subtitle as the main tittle hence I used >> to simply add and not replace.
To see the contents we have written so far we use cat README.md.
I also incorporated nano README.md to edit some parts where i saw fit to edit.

Initializing git

Then I used git init which is used to treat the folder as a git repository creating a hidden folder. To see the hidden files the command function ls -la is used.
I used the function git status to see the untracked files (these are files which exist but git hasn't been instructed to add them).

Committing and pushing

I then prepared changes for the next commit by using git add . followed by git status to confirm it is ready for commitment.
git commit -m "add Kenya Hospital Analysis Project" to save changes to the commit. The message in quotes indicates what changes I made.
Then I used git log --oneline to see the commit history.
A side note was the command function git branch -M main which would have been used if the output did not have main to it.
I further ran git remote add origin SSH where the SSH is on Github and its content is copy pasted directly after the origin.
Then used the command git remote -v to confirm the remote URL was added correctly and lastly git push -u origin main to upload it on Github.