Overview
Google Data Analytics Capstone: Complete a Case Study is the eighth and final required course. Instead of new technical content, it asks you to apply the whole program to one project. Track A offers two prepared scenarios, a bike-share company and a wellness technology company, and Track B lets you follow your own path with a public dataset you care about. Either way, the case study is organized by the six phases from Course 1, ending with recommendations a stakeholder could act on.
The course also covers what case studies and portfolios are for, how to host your work on platforms such as Kaggle, how to discuss it in interviews, and how to claim and share your certificate. Its final module adds AI skills: using generative AI to help with questions, cleaning, formulas and visualization ideas, while checking every output.
How it's tested
This course has no module challenges. It is assessed through completing the case study and short activities, including a knowledge check on professional case studies and an item confirming you completed a case study. Practice questions on this page use the same reasoning the capstone demands: which phase a step belongs to, what a section must contain and how to judge evidence.
-
Which section does this belong in?
A capstone task such as stating the business task, documenting cleaning steps or listing recommendations is described and you place it in the correct phase.
-
Is the evidence strong enough?
A finding and its supporting data are described and you decide whether the recommendation follows, or what caveat it needs.
-
Present the work
An interview or portfolio situation is described and you choose how to present the case study clearly and honestly.
A scenario says a learner's bike-share case study shows that casual riders take longer rides than annual members, and asks for the best act-phase recommendation. Candidates pick 'casual riders will convert if prices drop', which the data does not show. A stronger recommendation is tied to the evidence: target casual riders at the times and stations where long rides cluster, and test a membership offer there. Recommendations should follow from findings, not assumptions.
Modules 4
Official modules from the Google outline, in outline order.
8.1 Learn about capstone basicsWhat capstones, case studies and portfolios are and why employers value them.
- What it means
- A case study is a common way employers assess job skills by asking candidates to work through a realistic business problem; a portfolio is a collection of such work that shows your skills to employers.
- What to understand
- A strong case study states the business task, describes the data and how it was checked, documents cleaning, presents analysis and visuals, and ends with recommendations. Reviewing real analysts' portfolios shows the expected structure and level of polish, and a data journal helps you plan the project.
- How it shows up
- A description of a portfolio piece is given and you identify what is missing, such as a clear business question or recommendations.
- Common confusion
- Treating a portfolio as a list of tools used; employers look for problems solved and the reasoning behind each decision.
8.2 Optional: Build your portfolioChoosing a case study track and hosting the finished case study.
- What it means
- Track A provides Case Study 1 (a bike-share company) and Case Study 2 (a wellness company); Track B lets you follow your own case study path with a public dataset. The finished work is hosted on a platform such as Kaggle.
- What to understand
- Each case study is completed phase by phase: ask (business task and stakeholders), prepare (data sources, credibility, licensing and privacy), process (tools and cleaning documentation), analyze (summary and trends), share (visuals and key findings) and act (top recommendations and next steps). Track B demands extra care in selecting a dataset that passes ROCCC and is licensed for reuse.
- How it shows up
- A step in a case study is described and you place it in the correct phase or identify what the deliverable for that phase should include.
- Common confusion
- Skipping this module because it is labeled optional; it is where the case study itself is built.
8.3 Optional: Use your portfolioDiscussing your portfolio in interviews, elevator pitches and negotiation.
- What it means
- This module prepares you to present your case study to recruiters and interviewers, practice an elevator pitch, and negotiate terms after an offer.
- What to understand
- An elevator pitch is a short summary of who you are, what you do and what you want, and a portfolio walkthrough should explain the business problem, your approach, the result and what you would do differently. Interviewers often ask candidates to explain a decision in their analysis, so knowing why you cleaned or filtered data a certain way matters as much as the final chart.
- How it shows up
- An interview question about your case study is described and you choose the answer that explains reasoning and impact clearly.
- Common confusion
- Describing tools and steps without stating the business question and the outcome.
8.4 Put your certificate to workClaiming the certificate, job-search benefits and AI skills for data analytics.
- What it means
- The final module covers claiming and sharing your certificate and digital badge, the job-search resources that come with it, and an introduction to using AI in data analytics.
- What to understand
- Generative AI can suggest questions, help clean data, draft formulas and code, and propose visualization ideas, but its output can be wrong and must be verified against the data. Analysts should not paste confidential or personal data into AI tools without approval, and should be able to explain any AI-assisted step.
- How it shows up
- A scenario describes using AI to write a formula or clean data and asks what the analyst should do before relying on the result.
- Common confusion
- Treating AI-generated formulas or code as correct because they run without errors.
High-yield concepts
The terms, rules and formulas this domain turns into questions most often.
-
Case study
A realistic business problem worked through end to end to demonstrate skills, often used by employers in hiring.
- Remember
- It should read like a report to a stakeholder: question, data, method, findings, recommendations.
- Tested as
- The core deliverable of the course.
- Trap
- Submitting analysis with no stated business task or recommendation.
-
Portfolio
A public collection of projects, such as Kaggle notebooks, a GitHub repository or a personal site, that shows your work to employers.
- Remember
- One clear, well-documented case study is more persuasive than many unfinished ones.
- Tested as
- Portfolio modules and interviews.
- Trap
- Publishing code without explanation of decisions.
-
Business task statement
A clear statement of the problem the analysis must solve and who it serves, written in the ask phase.
- Remember
- It should be specific enough to judge whether the recommendations answer it.
- Tested as
- First deliverable in the case study.
- Trap
- Writing a topic, such as 'bike usage', instead of a question.
-
Prepare-phase deliverable
A description of the data sources, how they are organized, and whether they are credible, licensed and privacy-safe.
- Remember
- Apply ROCCC and note any bias or limitations before cleaning.
- Tested as
- Prepare section of the case study.
- Trap
- Skipping licensing and credibility notes for public datasets.
-
Process-phase documentation
A record of tools chosen and every cleaning or transformation step, such as a changelog.
- Remember
- Another analyst should be able to reproduce your cleaned dataset from it.
- Tested as
- Process section of the case study.
- Trap
- Presenting cleaned data without saying what changed.
-
Act-phase recommendations
Usually the top three evidence-based recommendations and next steps for the stakeholder.
- Remember
- Each recommendation should trace back to a specific finding.
- Tested as
- Final section of the case study.
- Trap
- Recommendations that rely on assumptions the data does not support.
-
Track A vs. Track B
Track A uses prepared case study scenarios; Track B lets you pick your own public dataset and question.
- Remember
- Track B adds responsibility for choosing a credible, licensed dataset and a clear business question.
- Tested as
- Planning the capstone.
- Trap
- Choosing Track B with a dataset that cannot answer an interesting question.
-
Elevator pitch
A short introduction covering who you are, what you do and what you want, delivered in under a minute.
- Remember
- Tie it to a concrete result from your case study.
- Tested as
- Interview preparation.
- Trap
- Listing tools rather than outcomes.
-
Responsible AI use
Using generative AI to support analysis while verifying outputs and protecting sensitive data.
- Remember
- Check AI-suggested formulas and code against known results before relying on them.
- Tested as
- Module 4 AI activities.
- Trap
- Pasting confidential data into a public AI tool.
-
Data journal
A planning document for the capstone that records decisions, questions and progress.
- Remember
- It doubles as raw material for documentation and interview answers.
- Tested as
- Module 1 activity.
- Trap
- Reconstructing decisions from memory at the end.
Side by side
Concepts candidates mix up, compared on the point that decides the answer.
Track A vs. Track B
| Concept | Meaning | Key difference | Common trap |
|---|---|---|---|
| Track A | Prepared case studies: a bike-share company or a wellness company | Business task and data are provided | Copying a common approach without your own reasoning |
| Track B | Your own case study path with a public dataset | You define the question and vet the data | Choosing a dataset that fails ROCCC or lacks a clear license |
Case study phase deliverables
| Concept | Meaning | Key difference | Common trap |
|---|---|---|---|
| Ask | Business task and key stakeholders | Defines what success means | Starting with charts instead of a question |
| Prepare and process | Data sources and credibility; cleaning steps and tools | Establishes that the data can be trusted | No record of cleaning decisions |
| Analyze and share | Summary of analysis; visuals and key findings | Turns data into insight for an audience | Charts without a stated takeaway |
| Act | Top recommendations and next steps | Converts findings into decisions | Recommendations the data does not support |
Portfolio piece vs. coursework submission
| Concept | Meaning | Key difference | Common trap |
|---|---|---|---|
| Portfolio piece | Public, polished case study for employers | Explains reasoning and impact for a reader who was not in the course | Assuming readers know the course context |
| Coursework submission | Work completed to meet course requirements | May skip explanation the grader already knows | Publishing it unedited as a portfolio piece |
Exam traps
Each trap is an answer that looks right. Learn the clue that should change your mind.
-
Charts as the conclusion
- What candidates pick
- End the case study with a dashboard of findings.
- Why it looks right
- Visuals look like a finished product.
- Why it's wrong
- The act phase requires recommendations and next steps; findings alone do not tell the stakeholder what to do.
Clue to watch for The scenario asks what the stakeholder should do next.
-
Recommendation beyond the data
- What candidates pick
- Recommend a price cut because it 'will convert' casual riders.
- Why it looks right
- It sounds decisive and commercially sensible.
- Why it's wrong
- The data shows ride behavior, not price sensitivity; recommendations must trace to evidence or be framed as tests.
Clue to watch for A recommendation that introduces a variable the analysis never measured.
-
Any public dataset will do
- What candidates pick
- A Track B dataset scraped from a site with no source or license.
- Why it looks right
- It matches an interesting topic.
- Why it's wrong
- Unverifiable or unlicensed data fails ROCCC and may not be reusable; choose a credible source with clear terms.
Clue to watch for No original source, citation or usage license is mentioned.
-
Undocumented cleaning
- What candidates pick
- Share only the final cleaned dataset and results.
- Why it looks right
- It keeps the write-up short.
- Why it's wrong
- Reviewers cannot trust or reproduce results without knowing what was changed; documenting cleaning is part of the process phase.
Clue to watch for The question asks what makes the case study credible to a reviewer.
-
AI output accepted as is
- What candidates pick
- Use an AI-suggested formula because it returns a number without errors.
- Why it looks right
- It saves time and appears to work.
- Why it's wrong
- AI tools can produce plausible but wrong logic; verify against a known subset or a manual calculation first.
Clue to watch for The scenario mentions using generative AI to write formulas, code or cleaning steps.
-
Tool list as the pitch
- What candidates pick
- Answer 'tell me about your project' by listing SQL, Tableau and Python.
- Why it looks right
- Tools are what job ads list.
- Why it's wrong
- Interviewers want the business problem, your approach, the result and what you learned; tools are supporting detail.
Clue to watch for An open-ended interview question about your portfolio.
Answer framework
Use these steps to plan and check your capstone case study before publishing it.
- Write the business task as a SMART question and name the stakeholders who will act on the answer.
- Describe your data sources, test them with ROCCC, and record licensing, privacy and known limitations. Note any gaps you will have to fix during cleaning.
- Clean the data with the right tool, logging every change and its reason so others can reproduce it.
- Analyze to answer the business task, then build only the visuals that support your key findings. Make sure each visual has a stated takeaway.
- Close with your top three recommendations, each traced to a finding, verify any AI-assisted steps, and publish with a short summary for employers.
Practice question
Want more like this? The Google Data Analytics practice test has questions for every domain.
Checklist
0 of 12
Tick each one you can do without notes. Your ticks stay in this browser.
Study strategy
Treat the capstone as a small project with a deadline. In Module 1, study two or three real analyst portfolios and note how they move from question to recommendation, then set up a data journal. Decide on a track early: Track A gives a defined business task and data, which suits most learners; choose Track B only if you have a credible, licensed dataset and a clear question.
Work the case study phase by phase, writing each section as you finish it rather than at the end. The most underestimated parts are the prepare and process write-ups: credibility, licensing and cleaning documentation are what make the work trustworthy to a reviewer. Pick the tools you know best from Courses 3 to 7; you can mix spreadsheets, SQL, Tableau and Python.
In the final module, try the AI activities on a copy of your data and check every suggestion. Afterward, publish the case study, rehearse a two-minute walkthrough and an elevator pitch, and take a mixed practice test across all courses to confirm your readiness.
FAQ
Is the Google Data Analytics capstone graded?
The capstone has no module challenges. Completion is tracked through short activities, including a knowledge check on professional case studies and an item confirming you completed a case study. The case study itself is mainly for your portfolio.
What case studies can I choose?
Track A offers Case Study 1, about a bike-share company, and Case Study 2, about a wellness company. Track B lets you follow your own path with a public dataset of your choice.
Why are two modules labeled optional?
The course page labels Build your portfolio and Use your portfolio as optional, but Build your portfolio is where the case study tracks are described and the work is hosted, so most learners complete it.
Which tools should I use for the capstone?
Any the program taught: spreadsheets, SQL in BigQuery, Tableau, and Python. Choose the tools that fit your data size and that you can explain confidently in an interview.
Can I use AI tools in my capstone?
The final module teaches using generative AI to help with questions, cleaning, formulas and visualization ideas. Verify every output and avoid sharing confidential or personal data with AI tools.
How long does the capstone take?
Coursera lists about 11 hours of course content. Our editorial planning estimate is 15 to 20 hours, though a polished Track B case study often takes longer.
Where should I publish my case study?
The course walks through Kaggle, and GitHub or a personal website also work. Include a short summary so a recruiter can grasp the business problem and your recommendations quickly.
Other Data & Analytics certifications
Where the same subject shows up in other exams. Cards marked "Same subject" open the matching domain.
Test your Domain 8 knowledge
You have covered planning, documenting and presenting a full case study, plus using AI responsibly. Check your readiness across all eight courses before you publish your capstone.
Keep going
- Training Training Built on your Weak areas
- Practice test Questions by domain, with explanations
- Mock exam Full length, timed like the real exam
- Readiness test Find your weakest domain in minutes
- Cheat sheet Key terms and formulas on one page
- Ask the Google Data Analytics AI tutor Get any concept explained your way