Analyst: Elene Kikalishvili
Date: November 2025
Tools Used: Excel Power Query (M Language), Google Sheets
AnalystForger is a global learning and mentorship platform that helps aspiring and transitioning professionals build real-world data analytics skills through guided projects, career-focused learning paths, and portfolio coaching.
As part of its Ultimate Project Builder initiative, AnalystForger conducted a global survey among data professionals and students to understand their experience levels, tool preferences, challenges, and career goals.
However, the raw data collected from Google Forms was messy and inconsistent - with irregular capitalization, multi-select responses stored in single cells, and missing regional classifications.
These inconsistencies made it difficult for the company to uncover meaningful insights or scale future analytics and curriculum-planning efforts.
To address this, I developed a fully automated data cleaning and preparation workflow using Excel Power Query (M Language).
The process dynamically connects to a Google Sheets source, cleans and standardizes all responses, enriches the dataset with regional mapping, and outputs structured, analysis-ready tables ready for Power BI or further exploration.
This workflow not only eliminated manual data preparation but also created a repeatable foundation for future surveys - allowing AnalystForger to:
- Analyze tool usage and learning goals at scale
- Understand learner challenges and motivations
- Identify preferred project domains
- Optimize survey engagement times across regions
By transforming unstructured survey data into a clean, enriched dataset, this project provides AnalystForger with a scalable, data-driven foundation for understanding its learners and improving program design.
The next step is to extend this workflow into automated dashboards and reporting, enabling continuous insight generation with minimal maintenance.
This project demonstrates a live data cleaning pipeline using Power Query. For privacy and reliability, the Google Sheets connection has been replaced with a sample dataset of identical structure. The final cleaned outputs are shared for full transparency.
AnalystForger collected responses from its global community of students and professionals to understand trends in data analytics learning. However, the data was messy and unstructured, requiring automation and enrichment.
Key objectives:
- Automate the import and cleaning process
- Standardize and categorize survey responses
- Add hierarchical geographic information (Region, Sub-Region, Continent)
- Derive engagement metrics by day and hour
- Structure data into normalized analytical tables
- Maintaine metadata and documentation for reproducibility
Deliver structured datasets that support insights into:
- Tool preferences (current proficiency and learning goals)
- Top requested project domains
- Common challenges in building data projects
- Respondents' career motivations and goals
- Participation patterns by region and time (day/hour)
The entire process was executed using Excel Power Query (M Language) with automated data retrieval from Google Sheets.
In-Scope Activities:
- Automated Data Import
- Data Cleaning & Standardization
- Data Transformation
- Geographical Enrichment
- Categorization & Grouping
- Timestamp Enrichment
- Multi-Selection Normalization
Out of Scope:
- Performing statistical or trend analysis
- Building dashboards or data visualizations
- Providing strategic recommendations
Supporting Documentation:
The following project deliverables are included in this repository:
- Automated Cleaning Workflow – Survey Data Cleaning Workbook
- Enriched Master Dataset – Survey Data Master Dataset
- Normalized Fact Tables – Five Fact Tables
- Power Query Script Documentation – Power Query Scripts
- Reference Dataset – UN Geo Mapping Dataset
- Data Cleaning Log – cleaning Log
- Statement of Work (SOW) – SOW
- Metadata / Data Description – Metadata
- Power Query (M Language)
- Data Cleaning & Transformation
- Data Normalization & Structuring
- Data Enrichment (Geo Mapping)
- Automated Data Refresh
- Documentation & Metadata Management
- Fully automated cleaning pipeline for dynamic Google Sheets survey data
- Enriched dataset with region, sub-region, and continent hierarchy
- Clean, structured tables for multi-selection questions
- Segmented respondents by role (Student vs Non-Student)
- Time-based insights through day/hour extraction
The cleaned dataset now allows AnalystForger to:
- Identify most popular tools and learning goals
- Understand common project challenges
- Discover preferred project domains
- Optimize survey scheduling and engagement strategy
- Import cleaned data into Power BI for visualization.
- Develop dashboards for tool, domain, and challenge insights.
- Fully automate refresh cycle via Power Automate or Python scripts.
- Extend enrichment with demographic or skill-level data.
⚠️ Disclaimer:
This project was created as a portfolio case study.
AnalystForger is a fictional organization created for educational and portfolio purposes.
The survey dataset used in this project was generated to simulate realistic responses and does not represent real individuals or proprietary data.