DEV Community: Faybeth Robina

KENYA CROP PERFORMANCE DASHBOARD

Faybeth Robina — Thu, 27 Nov 2025 20:49:27 +0000

Introduction.

In our country Kenya, the agriculture sector contributes greatly to our economy. Agriculture contributes significantly to exports earnings, food security and job
opportunities. Kenya produces variety of crops such as food crops and cash crops which include maize, beans, coffee, tea, potatoes. The agriculture sector undergoes different challenges which makes it necessary to monitor agricultural data. The Kenya crop production dashboard analyses the crop performance in different
counties in the country.

The dashboard is designed with the following sections;
The KPI cards.
The KPI cards shown are;
Total revenue, total profit, total yield and the total planted area.
Combo chart.
Line and stacked column chart where the columns show the seasons, sum of revenue of each county and the line chart shows the sum of profit earned from each
county.
Map chart.
The map displays the yield production level by county. It helps identify high and low producing counties. The bubbles appear big when the yield is high and small when the yield is low.
Column chart
The clustered column chart compares the total revenue and profit earned by crop type.
Line chart.

The line chart shows how revenue changed across the harvesting months. It also shows how profit increases and decreases over time.
Pie chart
Shows the percentage of yield produced by fertilizer used. It helps show how each fertilizer contributed to the yield production.
Interactive filters.
Filters used are by crop type, county and seasons. It helps to explore the dataset by the area one is interested in.
Insights.
Crops like rice, sorghum, potatoes, cassava and tea have high total revenue which shows that farmers earn high total earnings between the month of January and April.

When the season is long rains the sum of profit increase and counties like Nyeri, Nairobi, Kiambu, Kisumu and Nakuru get high revenue as shown below

During the long rains seasons counties like Machakos earn low profits and low revenue such regions need more support. Crops like potatoes and tomatoes don’t do well during this season too farmers are advised not to plant such crops during the long rain seasons.
Importance of the dashboard.
Helps to know which fertilizer to use when planting different crops to get more yield which in turns helps the farmer to earn more revenue and profits.

Helps researchers analyze agricultural trends to help farmers know where and which sectors to improve. Helps the policy makers in resource allocation and make polices that align the
region. Helps investors to identify areas to add their investments and during which periods.
Conclusion.
The dashboard makes agricultural data easy to understand. It helps compare regions and see trends which helps to improve farming in Kenya. Farmers can be helped with this dashboard to know where to make improvements to increase yield

Excel’s Contributions and Shortcomings in Predictive Analysis

Faybeth Robina — Thu, 27 Nov 2025 19:04:59 +0000

Microsoft Excel is a common spreadsheet program developed by Microsoft. Excel is designed for creating, organizing, analyzing and manipulating data in a tabular format. Excel is widely used in various businesses and industries because of its ability to handle different data related tasks and its versatility. Its powerful functionalities make it a tool for tasks ranging from simple calculations to complex data analysis and financial modeling. When it comes to predictive analysis and making data driven business decisions, Excel has both advantages and limitations.
This article breaks down Excel’s capabilities and limitations in predictive analysis and highlights its role in supporting strategic business decisions.

Excel’s strengths in predictive analysis.

Excel user-friendly interface makes it accessible to users with different level of expertise. Basic functions can be quickly learned and applied hence makes it easy to use.

Data analysis tools.
Excel has powerful tools for data analysis like formulas, functions, pivot tables and charts, helping in insightful data interpretation. It offers a range of charting options enabling one to represent data for better understanding and visual presentation.
The figure below shows the pivot table and a chart

Compatibility.
Excel files are compatible in that (.xlsx) are widely supported across various platforms, this ensures easy sharing and collaboration.
Logical formulars
Excel has custom calculations using functions like INDEX, MATCH, HLOOKUP, VLOOKUP, IF. These formulars are helpful for business modeling and estimating future trends based on historical data.
Integration with other tools.
Excel can integrate with tools like Power BI, Python, SQL or R allowing more advanced users to extend their skills into deeper analytics.

Excel’s weaknesses in predictive analysis

Limited data handling
Excel has limitations in handling large datasets efficiently, leading to performance issues and
potential data conflicts.
Prone to errors.
Human errors like incorrect formulars or data input can occur, leading to inaccurate results especially in complex spreadsheets. Excel relays on manual inputs, copy pasting and formular linking, this makes models prone to human error.

Statistics analysis limitation
Excel’s statistical functions are not comprehensive as those found in specialized software. Complex statistical analyses can be challenging and time consuming. Security risks, documentation and version control.
Excel spreadsheet lacks advanced security features like encryption, audit trails. Downloading data to spreadsheets can pose security risks. Excel lacks features for documenting assumptions,
tracking changes and managing different versions of a spreadsheet. This can make it difficult to track errors.
Regardless of its limitations, Excel can be used in making data driven business decisions.

The role of excel in making data driven business decisions.

Data cleaning and preparation.
Excel is used to identify and cleans, correct errors, handle missing values, removes duplicates and standardize formats which is helpful for preparing datasets for analysis. You can convert text to columns as shown below

Dashboard and reporting.
Excel dashboard and reporting helps in making data driven business decisions by providing a visual representation that is it condenses data into charts and graphs which offers a clear overview of business performance. It helps identify areas for improvement and make strategic decisions. The figure below shows an example of a dashboard on sales analysis

Conclusion.
Excel datasets can be integrated into other complex tools like Power BI to analyze data. Excel is a powerful tool for basic predictive analysis and data driven business decision making but for small businesses with medium or small datasets.

PHARMACEUTICAL SALES PERFORMANCE ANALYSIS.

Faybeth Robina — Wed, 15 Oct 2025 18:20:07 +0000

This report examines prescription (Rx) activity captured in the RCPA reporting form, benchmarking actual Rx against assigned brand targets, evaluating doctor conversion, and mapping competitive activity across regions.

Data Preparation & ETL (Power Query)

Import Data Sources

Load all required files: RCPA Reporting Form, PRODUCT MASTER, and BRAND TARGETS.
In Power BI, use Home > Get Data > Excel (or CSV/SQL) to import each dataset. RCPA & Competitor RCPA Transformation (Expected Transformation Sheet)
Go to Home > Transform Data to enter Power Query.
Merge/Append Queries: Combine datasets as needed (e.g., join product master to RCPA report by product or merge competitor info).

Data Cleanup and Transformation in Power Query

1.Splitting Tables Using Delimiters

Line Feed Split: Start by splitting rows where multiple products/quantities are listed together by a line feed (carriage return) delimiter. In Power Query:
Select the column to split. Use Home > Split Column > By Delimiter. Set delimiter to "Line feed" (often shown as #(lf) or with the Unicode line break symbol). This separates each product entry into individual rows.
Further Comma Split: After splitting by line feed, some columns (e.g., those with product names and quantities combined, such as "Aspiris-Pain,100") need to be split again—this time by a comma.
Select the target column. Split Column > By Delimiter, select comma (,). This gives separate columns for Product and Quantity.

2. Removing Duplicates

To ensure data integrity (especially when merging or appending), remove duplicates based on a combination of columns: UniqueID (often doctor or chemist ID, or visit ID). Product (to catch repeated entries for the same product from the same source).
In Power Query: With the cleaned-up table selected, Home > Remove Rows > Remove Duplicates.
Select or highlight both the UniqueID column and Product column before removing—the operation keeps only unique combinations of UniqueID and Product.

3. Additional Clean-up Steps (as needed)

Trim and Clean Text: Use Transform > Format > Trim and Clean to remove leading/trailing spaces and non-printable characters from all text columns.
Data Type Correction: Change columns (like Quantity) to numeric type for calculations. Filter Invalid Data:
Remove entries where critical columns (UniqueID, Product, Quantity) are blank or contain errors.

Build Summary Tables:

RCPA Data Table with Rx counts per doctor/product/region.
Competitor RCPA Table in similar structure.
Apply Changes: Click Close & Apply to commit ETL steps.

Building Visualizations

Doctor Rx Performance per Brand, Med Rep, and Region
Use Stacked Bar Chart or Matrix Visual.
Axis: Doctor/Med Rep/Region; Values: Rx Quantity; Legend: Brand.
Set up slicers for Doctor, Med Rep, Chemist, Product for interactive filtering.

Key performance indicator(KPIs)

Total prescription(Rx) Quantity(Qty)= 3,000, which shows the actual prescription volume achieved.
Total Target Rx Qty =74. This is the baseline target.
The performance rate is 37.78%.
The company maintains a strong competitive position with 3000 Rx volume while the competitor's 2000 Rx representing a 50% volume advantage .

Reginal Performance Analysis

Regional analysis reveals significant variation in prescription volumes across territories.

Top Performing Regions
Nyeri
Nyeri is the highest performing region. It has the highest volume market. The focus product prescription is 981 prescriptions while the competitors product prescription is 954 prescriptions. Leads by 5% margin in the highest volume market. This shows that the competitor is highly active.
North Rift
The focus product prescription is 798 prescriptions and the competitive product prescription is very low, this shows that it has maintained a strong performance .

Meru
Meru region has a mid tier performance. The focus product prescription is 621 prescriptions and the competitors is 432 prescriptions. There are 189 doctors, this shows excellent per doctor productivity. The performance is solid with a consistent competitive edge.
Underperforming Regions
Nairobi B
Nairobi B is one of the underperforming regions with the total prescription being 18 despite having 251 doctors. This represents a dramatic underperformance which requires urgent investigation.
Coast
The total prescription quantity is very low against the competitor's which is 248. Marketing the focus product in coast will help the volume of prescription to increase.
Doctors and medical representatives per region
The analysis of doctors and medical representatives distributions reveals a near one to one ratio across the regions.

Nairobi A
Nairobi A has 365 doctors and medical representatives generating
252 prescriptions. this indicates low efficiency.
Nairobi B
Nairobi B shows 251 doctors and medical representatives producing 18 prescriptions, representing very low efficiency.
Nyeri
Nyeri has 171 doctors and medical representatives generating 981 prescriptions, demonstrates high efficiency.
North Rift has 117 doctors and medical representatives generating 798 prescriptions also indicates very high efficiency.

In regions like Nairobi A and B we can see that it has the highest number of doctors and medical representatives but generates low prescriptions volumes, while Nyeri and North Rift with fewer doctors and medical representatives generate the highest volumes. This suggests that urban market saturations creates intense competition where multiple pharmaceutical companies compete for the same prescribers. Better market penetrations exist in mid sized cities, this indicates that these markets may be less saturated allowing for deeper relationships and more effective sales execution. The doctors quantity does not matter, the relationship quality matters. The substantial potential exists for resources reallocation from low yield regions to high yield regions this will improve the overall organizational performance.

Product Performance Analysis

Pie chart was used to show the total prescription per sales division and the number of doctors per sales division as shown below;

Across the three product divisions there exists a critical misalignment. The Aspiris Pain division covers 32.82% of doctors but generates 78.86% of prescription volume which earns an excellent performing rate. The Aspiris Gyn division engages 28.28% of doctors and produces 16.63% of prescription volume. The Aspiris Chronic division has 38.9% of doctors but only generating 4.51% of prescription volume.

The Pain division represents the primary revenue driver and the organization's flagship product. It yields an efficiency ratio of 2.40. With only a few doctors it generates four fifths of all prescription volume. This performance indicates strong product market fit, effective sales execution and competitive advantage. This division serves as the revenue engine and requires maximum protection and continued investment.

The Gyn division shows below average but acceptable performance. The efficiency ratio of 0.588 indicates room for improvement. Market potential clearly exists, with strategic investment in sales , targeted continuing medical education programs and improved marketing materials this division could move from moderate to strong performance.

The Chronic division represents a critical business problem demanding immediate executive attention. Having the largest doctor coverage yet generating minimal prescription volume indicates fundamental problems. The possible root causes include pricing concerns, prescribing barriers, training gaps, markets misalignments or competitive displacements. The organization is deploying 38.9% of its sales to achieve only 4.51% of results this represents massive resource waste.

Approval Status and Workflow Analysis

1623 prescriptions remain in progress status representing 58.05% of the total prescriptions, while 1173 prescriptions have achieved approved status of 41.95% .

Approximately 60% of prescriptions in pending status represents significant business risk. Revenue delay occurs as unapproved prescriptions do not generate immediate revenue. Prescription abandonment might occur as patients may seek alternatives during lengthy waits.

Converting 1623 pending prescriptions represents potential 145% revenue increase from the approved base. Process optimization would improve customer satisfaction. Fast approvals could benefit the organization.
Immediate actions to be taken are; implementing daily tracking of pending prescriptions, assigning dedicated teams to expedite approvals. Short term actions should be digitized approval processes, implement automated notifications. Long term initiatives should be develop predictive analytics and integrate healthcare providers systems.

JUMIA PRODUCT PERFORMANCE DASHBOARD ANALYSIS.

Faybeth Robina — Wed, 24 Sep 2025 17:21:24 +0000

Introduction
In our country Kenya Jumia is one of the largest online marketplace, offering a wide range of products like electronics, groceries, home appliance, home decor, beauty products and fashion products. It is part of the larger Jumia group, a pan African e-commerce company. Jumia Kenya was launched in May 2013 and overtime it has grown rapidly in terms of products, subscribers and vendors. Analyzing the dataset that contains 115 products, one can understand how discount, customers reviews and ratings influence the online shopping trends.

Dataset Overview
The raw dataset contained missing values, text based prices and inconsistent rating format. One has to follow the following steps to ensure the analysis is accurate :
Figure below, shows the raw data

Replaced blanks in the review and ratings columns with 0 for completeness.
Converted the prices columns to numeric format
Discounts converted from string to numerical values.
Ratings extracted from text format to numeric format (e.g. 4.5 out of 5 to 4.5)
Reviews standardized to numeric values.
Removed duplicates.

Data Enrichment

Added new columns to categorize products based on their ratings("Poor" for ratings below 3, "Average" for 3-4, and "Excellent" for 4.5 and above).
Created a new column to categorized products based on discount percentage ("Low Discount" for <20%, "Medium Discount" for 20-40%, "High Discount" for >40%).
Calculated the absolute discount amount for each product (Old Price - Current Price). Figure below shows the enriched data

It is important to clean your data because it produces a structured dataset ready for analysis.

Key Insights
The overall Key Performance Indicators(KPIs) are as follows;

Average price: Ksh 1,181
Average discount: approximately 36.8%
Highest discount: 64%
Lowest discount: 1%
Total reviews analyzed: 723
Average product rating: approximately 2

The figure below shows the Jumia performance dashboard.

Discount Analysis
some products with high discounts have very low reviews counts, this shows that price cuts alone don't guarantee engagement
The data shows that medium discount that is between 20% and 40% tend to attract more reviews than very high discounts.
Discounts appear effective when paired with higher ratings.

Ratings and Reviews Analysis
Products with higher ratings tend to have more reviews tend to have more reviews while poorly rated products have few to no reviews which indicates low engagement. Products with poor ratings have high discounts

Products
The bottom 5 lowest rated products are home decors, bags and they have high discounts and poor ratings. This might imply that the product quality has issues rather than pricing.
The figure below shows the top 5 low rated products.

The top 5 highest rated products have excellent ratings and medium to high discount rating this shows that the higher the ratings and reviews the higher the sale of the product.

Conclusion
In conclusion this report analysis shows us that discounting alone is not driving sustainable sales growth or customers loyalty. The analysis shows that while Jumia is offering a significant discounts the overall customer experience remains poor, the ratings and reviews are poor. Jumia needs to take different actions to ensure that there business is growing. Jumia should incentivize reviews by offering voucher rewards, loyalty points, or discount coupons for verified feedback.
Jumia must shift from price-driven competition to a value-driven strategy. By elevating product quality, spotlighting high-rated categories, optimizing discount ranges, and building trust through reviews, Jumia can boost conversions, reduce returns, and grow long-term customer loyalty.

JUMIA PRODUCT PERFORMANCE DASHBOARD ANALYSIS.

Faybeth Robina — Mon, 22 Sep 2025 21:11:41 +0000

Replaced blanks in the review and ratings columns with 0 for completeness.
Converted the prices columns to numeric format
Discounts converted from string to numerical values.
Ratings extracted from text format to numeric format (e.g. 4.5 out of 5 to 4.5)
Reviews standardized to numeric values.
Removed duplicates.

Data Enrichment

Added new columns to categorize products based on their ratings("Poor" for ratings below 3, "Average" for 3-4, and "Excellent" for 4.5 and above).
Created a new column to categorized products based on discount percentage ("Low Discount" for <20%, "Medium Discount" for 20-40%, "High Discount" for >40%).
Calculated the absolute discount amount for each product (Old Price - Current Price). Figure below shows the enriched data

It is important to clean your data because it produces a structured dataset ready for analysis.

Key Insights
The overall Key Performance Indicators(KPIs) are as follows;

Average price: Ksh 1,181
Average discount: approximately 36.8%
Highest discount: 64%
Lowest discount: 1%
Total reviews analyzed: 723
Average product rating: approximately 2

The figure below shows the Jumia performance dashboard.

The top 5 highest rated products have excellent ratings and medium to high discount rating this shows that the higher the ratings and reviews the higher the sale of the product.

STEP BY STEP GUIDE TO POSTGRESQL INSTALLATION AND SETUP USING A LINUX SERVER.

Faybeth Robina — Sun, 03 Aug 2025 12:53:12 +0000

INTRODUCTION
PostgreSQL, frequently abbreviated as Postgres, is one of the most advanced open-source relational database management system available these days. It is world wide known and used because of its support for advanced data type and extensibility. With internet connection, terminal access, simple commands and a linux system one will be able to install and set up Postgres on a linux server. This article will guide you through each step to ensure your system is set up correctly for PostgreSQL usage.

Step 1:Update System Package.
Ensuring your system package is up to date is important before installation, this will help to prevent potential conflict and also help you get the latest version of PostgreSQL. You can update the system package with the following command:
sudo apt update

Step 2: Installing PostgreSQL.
The next step is to install PostgreSQL and its dependencies. You will use the command below:
sudo apt install postgresql postgresql-contrib

Step 3: Initializing the Database.
The database cluster needs to be initialized. After the installation is complete PostgreSQL service will start automatically.

Step 4:Starting the PostgreSQL Service.
After initialization, you need to check the status of PostgreSQL service using the command :
sudo system status postgresql
If you do not see "active (running)" in green then input the command:
sudo systemctl start postgresql

Step 5: Verifying the Installation.
To ensure that the PostgreSQL is installed correctly you verify the server status as shown below.

Step 6: Accessing PostgreSQL.
During installation PostgreSQL creates a system user called postgres. This user can be used to manage the database. You switch to the postgres user with:
sudo -i -u postgres

as shown above once switched, access the PostgreSQL prompt with:
psql

Step 7: Create a New User and Database
To create a new user and database you can run the following commands:

CREATE USER your_username WITH PASSWORD'your_password';
CREATE DATABASE your_database_name;
GRANT ALL PRIVILEGES ON DATABASE your_database_name TO your_username;

Step 8: Configuring PostgreSQL to Allow Remote Connections
PostgreSQL is configured to listen only on the localhost. To allow remote connections you modify the configuration in the postgresql. conf file and the pg_hba.conf file. First, open the postgresql. conf file with the command sudo nano/etc/postgresql/<version>/main/postgresql.conf, then edit the line # listen_addresses = 'localhost' to listen_addresses = '*'. This change allows PostgreSQL to listen on all IP addresses. Next, edit the pg_hba.conf file to add a line that permits connections from remote IP adressses.

sudo nano/etc/postgresql /<version>/main/pg_hba.conf
       host    all     all  
   0.0.0.0/0           md5

These configurations allow all users from any IP address to connect to all database. Finally ,restart PostgreSQL to apply the changes by running the code:
sudo systemctl restart postgresql

You can now start using PostgreSQL on your Ubuntu system. You can connect to the PostgreSQL database using variety of tools like PgAdmin or DBeaver.

The image above shows how to connect PostgreSQL database to DBeaver where
Host: your server's IP address
Port: 5432
Username: your database user your_username
Password: password you created (your_password)
Database:your_database_name

Conclusion
Following the steps above you can install and set up PostgreSQL on your linux server. You can start utilizing its powerful features for your database needs. PostgreSQL can be used by both small and large scale applications.