8 Best First Analytics Projects for Career Changers

Data analytics is one of the fastest-growing career paths in the United States. The Bureau of Labor Statistics projects 34% growth for data scientist roles through 2034 — more than ten times the average for all occupations. Entry-level data analysts earn $63,000 to $88,000 annually, with mid-level roles reaching $93,000 (Glassdoor, 2026). But employers do not hire based on salary potential. They hire based on proof that you can work with data.

According to hiring data from InterviewQuery and Burtch Works, 72% of hiring managers say a portfolio matters more than a certification when evaluating entry-level analytics candidates. That means your first projects carry outsized weight in the hiring decision. Traecta — Your Personalized Career Roadmap maps your existing skills to the right projects so you build proof, not just follow tutorials. If you are starting from scratch, this learning path for adults changing careers into tech helps you build a study plan that fits your schedule.

Here are eight analytics projects that career changers can complete in one to four weeks each, using free tools and publicly available data.

What makes a good first analytics project#

Before jumping into projects, understand what hiring managers actually evaluate. Not all projects are equal in a portfolio. The ones that get attention share four traits:

1. Real data, not synthetic: Datasets from Kaggle, government portals, or public APIs signal that you can handle the messiness of real-world information
2. Clear business question: "What drives customer churn?" is a project. "Here is a pivot table" is an exercise
3. Written narrative: A dashboard without explanation is decoration. A dashboard with a written analysis is proof of communication skill
4. End-to-end process: Cleaning, analyzing, visualizing, and interpreting — not just one step

If you already compared certificates versus portfolio proof, you know the logic. Projects are what turn learning into evidence. For those moving from operations, finance, or HR specifically, the best career path from operations to analytics shows how to leverage your domain knowledge in each project.

Project comparison at a glance#

#	Project	Tools	Dataset Source	Estimated Time	Skill Focus
1	Customer churn analysis	Excel / Python / Tableau Public	Kaggle: Telco Churn (7,043 rows)	1-2 weeks	Cleaning, statistical analysis, visualization
2	Sales performance dashboard	Excel → Tableau Public	Kaggle: Superstore Sales (9,800 rows)	2-3 weeks	Dashboard design, KPIs, trend analysis
3	Survey data analysis	Google Sheets / Python	Pew Research Center or Kaggle	1-2 weeks	Survey methodology, cross-tabulation
4	Marketing campaign analysis	Excel / Tableau Public	Google Analytics Sample Dataset	2-3 weeks	Funnel analysis, cohort metrics
5	A/B test analysis	Google Sheets / Python	Kaggle: AB Testing datasets	1-2 weeks	Statistical significance, hypothesis testing
6	Employee attrition analysis	Excel / Python / Tableau Public	Kaggle: IBM HR Analytics (1,470 rows)	1-2 weeks	HR metrics, classification analysis
7	Financial analysis project	Excel / Python	Yahoo Finance or FRED	2-3 weeks	Time series, ratio analysis, benchmarking
8	Personal spending tracker	Google Sheets / Python	Your own data	2-4 weeks	Data collection, budget analysis, visualization

Project 1: Customer churn analysis#

What you will build: An analysis identifying which factors predict customer churn — the point where a customer stops using a service. You will explore patterns in demographics, contract types, and usage behavior to find the strongest churn drivers.

Recommended tools: Google Sheets or Excel for tabular analysis; Tableau Public for visualization; Python (pandas, matplotlib) if you want to add statistical analysis

Dataset: IBM Telco Customer Churn on Kaggle — 7,043 customer records with 21 features including tenure, monthly charges, contract type, and churn status. One of the most widely-used beginner datasets in analytics.

Estimated time: 1-2 weeks (20-40 hours)

What it proves to employers: That you can explore a real business dataset, identify patterns, and communicate findings. Churn analysis is a core task in marketing analytics, product analytics, and customer success roles — it appears in job postings across industries.

Step-by-step approach:

1. Download the dataset and inspect the structure — check for missing values, data types, and inconsistent formatting
2. Calculate the overall churn rate and break it down by contract type, tenure group, and payment method
3. Build a churn rate comparison table showing which segments have the highest and lowest churn
4. Create 3-5 visualizations: churn by tenure, churn by monthly charge range, churn by contract type
5. Write a one-page summary with three actionable recommendations for reducing churn

Common mistakes to avoid:

• Skipping data cleaning — the dataset has missing values in TotalCharges and inconsistent formatting. Handling these is part of the project
• Building a machine learning model when the task calls for descriptive analytics. Employers hiring entry-level analysts want to see clear explanations, not black-box predictions
• Presenting charts without written interpretation. Every visualization should answer a specific question

Project 2: Sales performance dashboard#

What you will build: An interactive dashboard tracking sales performance across product categories, regions, and time periods. The dashboard should include KPIs (total revenue, average order value, top products) and allow filtering by time range or region.

Recommended tools: Excel for initial analysis, then rebuild in Tableau Public (free, up to 15 million rows per workbook). Tableau Public is the standard for analytics portfolios because hiring managers can view your dashboard online without downloading files.

Dataset: Superstore Sales on Kaggle or any retail sales dataset with date, category, region, and revenue columns. The key is having enough rows (1,000+) and enough dimensions for meaningful slicing.

Estimated time: 2-3 weeks (30-50 hours)

What it proves to employers: Dashboard design is one of the most requested skills in data analyst job postings. According to the Stack Overflow Developer Survey, SQL and data visualization tools consistently rank among the top in-demand technical skills. This project demonstrates that you can move from raw data to a polished, interactive presentation.

Step-by-step approach:

1. Clean the dataset — handle returns, missing values, and date formatting
2. Calculate KPIs: total revenue, profit margin, average order value, year-over-year growth
3. Build a simple Excel dashboard first to validate your calculations
4. Rebuild in Tableau Public with interactive filters and drill-down capability
5. Publish to Tableau Public and link from your portfolio

Common mistakes to avoid:

• Using too many colors or chart types. A good dashboard uses 3-4 chart types maximum
• Forgetting the written summary. Include a brief explanation of your design choices and key findings
• Skipping the Excel validation step. Building in Excel first catches calculation errors before they become embedded in the dashboard

Project 3: Survey data analysis#

What you will build: An analysis of public survey data that tests hypotheses about attitudes, behaviors, or trends. You will demonstrate skill in working with survey methodology — weighting, cross-tabulation, and handling response bias.

Recommended tools: Google Sheets for basic analysis; Python (pandas, scipy) for statistical testing; Tableau Public for visualization

Dataset: Pew Research Center publishes downloadable datasets on technology use, workforce trends, and social attitudes. Kaggle also hosts survey datasets on topics from mental health to remote work preferences. Choose a survey relevant to your target industry.

Estimated time: 1-2 weeks (15-30 hours)

What it proves to employers: Survey analysis is a staple skill in market research, UX research, HR analytics, and public policy roles. This project shows you can handle categorical data, calculate meaningful proportions, and avoid common pitfalls like leading questions or sampling bias.

Step-by-step approach:

1. Read the survey methodology documentation — understand the sample size, demographics, and any weighting applied
2. Select 2-3 hypotheses to test: "Remote workers report higher job satisfaction than in-office workers," for example
3. Calculate cross-tabulations and conditional proportions
4. Run chi-square tests or t-tests (if using Python) to check statistical significance
5. Write a two-page report with methodology, findings, and limitations

Common mistakes to avoid:

• Ignoring the methodology documentation. Survey data requires understanding who was surveyed and how questions were worded
• Treating correlation as causation in your written conclusions
• Skipping the limitations section. Acknowledging what the data cannot prove shows analytical maturity

Project 4: Marketing campaign analysis#

What you will build: An analysis of website or marketing campaign performance using behavioral data. You will calculate conversion funnels, traffic sources, and engagement metrics — the same work marketing analysts do daily.

Recommended tools: Google Sheets or Excel for data processing; Tableau Public for visualization; Google BigQuery (free tier) if using the Google Analytics sample dataset directly

Dataset: Google Analytics Sample Dataset — obfuscated data from the Google Merchandise Store, available through Google BigQuery public datasets and Kaggle. Contains session data, traffic sources, product views, and transactions.

Estimated time: 2-3 weeks (25-45 hours)

What it proves to employers: Marketing analytics is a high-demand specialization. Understanding funnel analysis, cohort metrics, and attribution shows you can connect data to business outcomes. This project is especially strong for career changers coming from marketing, sales, or operations backgrounds.

Step-by-step approach:

1. Explore the dataset structure — sessions, users, pageviews, transactions
2. Build a conversion funnel: landing page → product view → add to cart → purchase
3. Analyze traffic sources: organic search, paid ads, direct, referral — which drives the highest conversion rate?
4. Calculate cohort-level metrics: new vs. returning users, average sessions per user
5. Create a dashboard with key findings and three optimization recommendations

Common mistakes to avoid:

• Reporting vanity metrics (pageviews, sessions) without connecting them to business outcomes (conversions, revenue)
• Not accounting for seasonality in time-based analysis
• Overcomplicating the analysis when a clear funnel and source breakdown tells the story

Project 5: A/B test analysis#

What you will build: A statistical analysis of an A/B test comparing two versions of a product, page, or campaign. You will formulate a hypothesis, check for statistical significance, and write a recommendation — the same process product analysts and growth teams follow.

Recommended tools: Google Sheets for basic calculations; Python (scipy.stats, statsmodels) for rigorous testing; a calculator tool like AB Test Calculator for quick validation

Dataset: AB Testing Practice Dataset on Kaggle or any dataset with control and treatment groups and a binary outcome. You can also use NeurIPS curated A/B test datasets for academic rigor.

Estimated time: 1-2 weeks (15-25 hours)

What it proves to employers: A/B testing is a core skill in product analytics, marketing analytics, and growth roles. This project demonstrates that you understand experimental design, statistical significance, and evidence-based decision-making — skills that differentiate analysts from report builders.

Step-by-step approach:

1. State the hypothesis: "Version B increases conversion rate compared to Version A"
2. Calculate the conversion rate for both control and treatment groups
3. Run a two-proportion z-test to check if the difference is statistically significant
4. Calculate practical significance: is the lift large enough to matter for the business?
5. Write a one-page decision memo with recommendation, confidence level, and potential next steps

Common mistakes to avoid:

• Stopping analysis at "p-value < 0.05" without discussing practical significance
• Not checking for sample size adequacy before running the test
• Confusing statistical significance with business impact. A 0.5% lift might be statistically significant but commercially irrelevant

Project 6: Employee attrition analysis#

What you will build: An analysis of what drives employee turnover using HR data. You will identify risk factors, calculate attrition rates by department and role, and recommend retention strategies — work that HR analysts and people analytics teams perform regularly.

Recommended tools: Excel for basic analysis; Python (pandas, seaborn) for deeper exploration; Tableau Public for a dashboard showing attrition patterns

Dataset: IBM HR Analytics Employee Attrition & Performance on Kaggle — 1,470 employee records with 35 features including age, department, job satisfaction, overtime status, and attrition label. A clean, structured dataset ideal for career changers.

Estimated time: 1-2 weeks (20-35 hours)

What it proves to employers: This project is particularly powerful for career changers coming from HR, operations, or management backgrounds. It demonstrates that you can combine domain knowledge with analytical methods. For operations professionals transitioning to analytics, this project connects directly to your existing experience with workforce management.

Step-by-step approach:

1. Calculate overall attrition rate and compare across departments, job roles, and tenure groups
2. Analyze the relationship between attrition and factors like overtime, job satisfaction, distance from home, and income
3. Build a risk profile: which employee segments have the highest attrition probability?
4. Create visualizations: attrition by department bar chart, income distribution for leavers vs. stayers
5. Write three specific, actionable retention recommendations based on data

Common mistakes to avoid:

• Treating 1,470 records as a large dataset. It is small enough to analyze thoroughly — use that advantage
• Ignoring confounding variables. Income and job level correlate with both tenure and attrition — acknowledge these relationships
• Making generic recommendations ("improve employee satisfaction") instead of data-specific ones ("employees working overtime in the Research department have 2.3x higher attrition — target them for workload review")

Project 7: Financial analysis project#

What you will build: A time-series analysis of stock prices, company financials, or economic indicators. You will calculate returns, volatility, and comparative performance — work that financial analysts and FP&A teams perform regularly.

Recommended tools: Excel for ratio analysis; Python (pandas, yfinance) for time-series data; Google Sheets for budget analysis if using personal or sample budget data

Dataset: Yahoo Finance data via Python yfinance library (free, real market data); Federal Reserve Economic Data (FRED) for macroeconomic indicators; or any company financial dataset from Kaggle.

Estimated time: 2-3 weeks (30-50 hours)

What it proves to employers: Financial analysis is valuable across industries — not just in finance. Every company needs analysts who can read financial statements, track budget performance, and benchmark metrics. This project is especially relevant for career changers from accounting, finance, or business operations.

Step-by-step approach:

1. Select 3-5 companies or assets and download historical price data (2-5 years)
2. Calculate key metrics: daily returns, cumulative returns, volatility, moving averages
3. Build a comparative performance table and benchmark against a market index
4. Create a time-series visualization showing price trends and volatility
5. Write an investment summary or budget analysis report with clear conclusions

Common mistakes to avoid:

• Making investment recommendations without disclaimers. This is a portfolio project, not financial advice
• Using only line charts. Financial analysis benefits from candlestick patterns, distribution plots, and correlation matrices
• Ignoring dividends and total return. Price appreciation alone tells an incomplete story

Project 8: Personal spending tracker with insights#

What you will build: A data collection and analysis project based on your own spending data. You will design a tracking system, clean your transaction data, categorize expenses, and surface actionable insights about your spending patterns.

Recommended tools: Google Sheets (easiest for data entry); Python (pandas) for analysis; Tableau Public or Google Sheets charts for visualization. You can export bank transaction data as CSV from most banking apps.

Dataset: Your own transaction data (exported from your bank, credit card, or budgeting app). If you prefer not to use personal data, use Kaggle consumer spending datasets as a substitute.

Estimated time: 2-4 weeks (20-60 hours, depending on how long you track)

What it proves to employers: This project demonstrates the complete analytics workflow end-to-end — from designing a data collection process through cleaning, analysis, and communication. It shows initiative and self-direction, which hiring managers value especially in career changers. Building job-ready portfolio projects covers the case study structure you should use when presenting this project.

Step-by-step approach:

1. Define categories: housing, food, transportation, subscriptions, entertainment, savings
2. Track spending for at least 30 days — enter transactions as they happen or export from banking apps weekly
3. Clean the data: remove duplicates, standardize merchant names, assign categories
4. Analyze: calculate monthly totals, category breakdowns, week-over-week trends, average daily spend
5. Build a dashboard and write a "spending audit" with three specific adjustments you plan to make

Common mistakes to avoid:

• Tracking for fewer than 30 days. One or two weeks of data produces unreliable patterns
• Over-engineering the tracking system. A simple Google Sheet with date, amount, category, and notes is sufficient
• Not following through with the insights step. The analysis is the project — not the spreadsheet itself

How to present your analytics portfolio#

The projects themselves matter, but presentation matters just as much. Hiring managers review dozens of portfolios. Make yours easy to scan.

Portfolio structure that hiring managers can navigate#

Use this structure for every project page in your portfolio:

Section	What to include
Problem	What business question did you answer?
Data	Source, size, and any cleaning steps
Method	Tools used and analysis approach
Findings	Key metrics and visualizations
Recommendations	Actionable next steps based on data
Reflection	What you would improve with more time or data

This structure mirrors how analysts present work in real organizations. If you are still deciding which role to target, your data analyst roadmap for experienced professionals helps you align projects with specific job requirements.

Platform options for hosting your portfolio#

Platform	Cost	Best for
GitHub Pages	Free	Projects with code (Python, SQL files, Jupyter notebooks)
Tableau Public	Free	Dashboards and visual analyses
Google Sites	Free	Simple project pages with embedded images
Notion	Free tier	Written case studies with embedded visuals

Use a combination. Host code on GitHub, dashboards on Tableau Public, and written analyses on a simple site that links everything together.

Picking the right projects for your background#

Not all eight projects are equally relevant for every career changer. Your existing professional experience should guide your selection.

Your background	Best starting projects	Why
Marketing, communications	Projects 2, 4, 5	Dashboard skills, funnel analysis, and A/B testing connect directly to marketing analytics roles
HR, people operations	Projects 6, 3	Employee data analysis and survey methodology match people analytics and HR reporting roles
Finance, accounting	Projects 7, 2	Financial analysis and sales dashboards align with FP&A and business analytics positions
Operations, logistics	Projects 1, 6, 8	Churn analysis and operational tracking map to supply chain analytics and operations analyst roles
Teaching, education	Projects 3, 5	Survey analysis and statistical testing leverage your understanding of assessment and methodology
Customer service, support	Projects 1, 4	Customer churn and campaign analysis build on your experience with customer behavior

The strongest portfolios mix one project closely tied to your domain and one project that stretches into a new area. This combination shows depth in your field and versatility across problems. Building a learning plan based on transferable skills can help you identify which skills from your current role translate directly to analytics work.

Tools summary: everything you need is free#

Tool	Cost	What it is for
Google Sheets	Free	Data entry, basic analysis, simple visualizations
Microsoft Excel	Often available through work/school	Pivot tables, formulas, basic dashboards
Tableau Public	Free (up to 15M rows)	Interactive dashboards, published online
Python (Google Colab)	Free	Data cleaning, statistical analysis, automation
SQL (DB Browser for SQLite)	Free	Database querying practice
Google BigQuery	Free tier (1TB/month)	Large-scale SQL queries on public datasets

Start with the tools you already know. Add new tools as each project requires them. There is no value in learning five tools simultaneously when your first two projects only need two.

Your first analytics project does not need to be impressive. It needs to be complete — with a question, a method, findings, and a written conclusion. One finished project that demonstrates real analytical thinking is worth more than three half-completed tutorials. If you want a structured path from where you are to your first analytics role, Traecta — Your Personalized Career Roadmap identifies your skill gaps and builds a project plan around your existing experience.