Overview
Ask Questions to Make Data-Driven Decisions is the second required course and covers the first phase of the analysis process. It teaches how to write effective questions using the SMART framework (specific, measurable, action-oriented, relevant, time-bound), how to classify a business problem into one of six problem types, and how data, reports and dashboards support decisions. Its longest module, Spreadsheet magic, covers formulas, functions, cell references, common error codes and the scope of work that structured thinking produces.
The final module turns to people: identifying stakeholders, managing their expectations, communicating clearly and handling conflict. Together these skills decide whether an analysis answers the question the business actually has. Courses 3 to 6 assume the question is already well defined, so a weak ask phase becomes wasted effort later.
How it's tested
Module challenges in this course mix classification items (which problem type, which SMART element is missing) with practical spreadsheet items (what a formula returns, what an error code means, which reference to lock). Stakeholder items are short workplace scenarios asking what the analyst should do first or say next. The right answer usually clarifies before acting.
-
Fix the question
A stakeholder's question is shown and you pick the rewrite that is specific, measurable, action-oriented, relevant and time-bound, or identify why the original is leading, closed or vague.
-
Classify the problem
A business need such as predicting next quarter's demand or flagging an unusual spike in returns is described and you name which of the six problem types it is.
-
Read the spreadsheet
A formula, function or error code is shown and you identify the result, the cause of the error or the correct cell reference.
A manager asks, 'Don't you think our new loyalty app is improving sales?' The question asks for the best SMART rewrite. The original is leading and closed, inviting a yes. The strongest option names the metric and timeframe without assuming the answer: 'How did average monthly spend change among app users versus non-users in the six months after launch?' It is specific, measurable, relevant and time-bound, and the result can drive action.
Modules 4
Official modules from the Google outline, in outline order.
2.1 Ask effective questionsSMART questions, the six problem types and questions to avoid.
- What it means
- Effective questions follow SMART: specific, measurable, action-oriented, relevant and time-bound. Google sorts business problems into six types: making predictions, categorizing things, spotting something unusual, identifying themes, discovering connections and finding patterns.
- What to understand
- Leading questions suggest their own answer, closed-ended questions invite only yes or no, and vague questions lack context; all three fail SMART. Categorizing assigns items to defined groups, identifying themes groups categories into broader concepts, discovering connections finds links between entities, and finding patterns uses historical data to see what happened over time.
- How it shows up
- A business request is described and you choose the problem type, or a question is shown and you pick the SMART rewrite.
- Common confusion
- Mixing up categorizing things with identifying themes; categorizing assigns individual items to labels, while themes group those categories into higher-level ideas.
2.2 Make data-driven decisionsHow data shapes decisions, quantitative versus qualitative data, and reports versus dashboards.
- What it means
- Data-driven decision-making uses facts to guide strategy, while data-inspired decision-making explores data for ideas without relying on it alone. Quantitative data is measured and counted; qualitative data describes qualities and reasons.
- What to understand
- A report is a static, high-level snapshot of data at a point in time that is easy to produce and share; a dashboard monitors live, incoming data in one interactive place but takes more effort to build and maintain. Metrics are single quantifiable measures used for comparison, and the course also contrasts small data (specific, short-term) with big data (large, less specific, long-term).
- How it shows up
- A stakeholder needs either a monthly snapshot or continuous monitoring, and you choose a report or a dashboard; or you identify whether data is quantitative or qualitative.
- Common confusion
- Assuming a dashboard is always better; a report is the right choice when the data is a fixed snapshot and stakeholders need something simple.
2.3 Spreadsheet magicFormulas, functions, cell references, error codes, scope of work and structured thinking.
- What it means
- Spreadsheets support every phase of the analysis process. This module teaches formulas and functions, relative versus absolute references, how to read error messages, and how structured thinking produces a scope of work.
- What to understand
- An absolute reference such as $B$2 stays fixed when copied, while a relative reference such as B2 shifts. #DIV/0! means division by zero or a blank cell, #REF! means a referenced cell no longer exists, #NAME? means an unrecognized function name, #VALUE! means the wrong data type, and #N/A means a lookup value was not found. A scope of work lists deliverables, timeline, milestones and reports.
- How it shows up
- A formula is copied down a column and returns wrong results or an error, and you identify the cause or fix.
- Common confusion
- Treating #N/A and #REF! as the same; #N/A means a value was not found, while #REF! means the formula points to a cell that was deleted.
2.4 Always remember the stakeholderManaging expectations, clear communication and teamwork.
- What it means
- Stakeholders are people who invest time and resources in a project and care about its outcome, such as executives, customer-facing teams and data science teams. Analysts must understand what each needs and when.
- What to understand
- Clear communication starts with four questions: who your audience is, what they already know, what they need to know and how to communicate it effectively. When requests conflict or expectations are unrealistic, the analyst clarifies the goal, explains the trade-off between speed and accuracy, and keeps stakeholders informed rather than silently changing scope.
- How it shows up
- A stakeholder asks for results sooner than the data allows, or two stakeholders want different things, and you choose the best first response.
- Common confusion
- Treating 'deliver whatever was asked, as fast as possible' as good service; the course prefers clarifying the real need and being transparent about limits.
High-yield concepts
The terms, rules and formulas this domain turns into questions most often.
-
SMART questions
Questions that are specific, measurable, action-oriented, relevant and time-bound, such as 'What share of trial users upgraded within 30 days in Q2?'
- Remember
- In Google's version, 'A' stands for action-oriented: the answer should support a change or decision.
- Tested as
- The most tested idea in Module 1.
- Trap
- Using the goal-setting version where A means achievable.
-
Six problem types
Making predictions, categorizing things, spotting something unusual, identifying themes, discovering connections and finding patterns.
- Remember
- Predictions use data to forecast outcomes; finding patterns uses historical data to understand what happened.
- Tested as
- Classification items in Module 1 and the capstone ask phase.
- Trap
- Labeling every trend question 'making predictions' even when it only looks backward.
-
Leading and closed-ended questions
A leading question pushes toward an answer; a closed-ended question allows only a short or yes-or-no reply.
- Remember
- Both reduce the information you get and can bias results.
- Tested as
- 'Which question should the analyst avoid' items.
- Trap
- Thinking a short, specific yes-or-no question is SMART just because it is specific.
-
Data-driven vs. data-inspired decisions
Data-driven decisions rest on facts from data; data-inspired decisions use data to explore ideas while drawing on other inputs.
- Remember
- Data-driven still needs context; data alone does not explain why something happened.
- Tested as
- Definition and scenario items in Module 2.
- Trap
- Treating them as identical terms.
-
Report vs. dashboard
A report is a static snapshot; a dashboard tracks live data interactively in one place.
- Remember
- Reports are quick to make but go stale; dashboards stay current but need maintenance and can be misread without context.
- Tested as
- Tool-choice items in Module 2.
- Trap
- Choosing a dashboard for a one-time board summary.
-
Metric and metric goal
A metric is a single quantifiable measure used for evaluation, such as conversion rate; a metric goal is the measurable target for it.
- Remember
- Return on investment (ROI) is net profit over a period divided by the cost of investment.
- Tested as
- Measurable questions and dashboard design.
- Trap
- Calling a raw data point, such as one day's sales, a metric.
-
Absolute vs. relative reference
$A$1 stays fixed when copied; A1 adjusts relative to its new position.
- Remember
- Lock the cell that holds a constant, such as a tax rate, before copying a formula.
- Tested as
- Formula-copy questions in Module 3.
- Trap
- Forgetting the $ and getting wrong values or #REF! after copying.
-
Spreadsheet error codes
#DIV/0!, #N/A, #NAME?, #NUM!, #REF!, #VALUE! and, in Google Sheets, #ERROR! each signal a different problem.
- Remember
- #NAME? usually means a misspelled function; #VALUE! means text where a number is expected.
- Tested as
- Error-diagnosis items in Module 3.
- Trap
- Guessing from the first letter rather than knowing each code's cause.
-
Structured thinking and scope of work
Structured thinking recognizes the problem, organizes information, finds gaps and identifies options; a scope of work (SOW) records deliverables, timeline, milestones and reports.
- Remember
- The SOW is agreed before analysis to prevent scope creep.
- Tested as
- Module 3 project-planning items.
- Trap
- Confusing an SOW with a statement of the findings.
-
Stakeholder communication questions
Who is the audience, what do they already know, what do they need to know, and how can you communicate effectively?
- Remember
- Tailor depth and detail to the audience; executives need the answer and its impact first.
- Tested as
- Module 4 scenario items.
- Trap
- Sending the full technical workbook to an executive audience.
-
Speed vs. accuracy trade-off
Stakeholders often want results fast, but rushing can produce wrong conclusions.
- Remember
- Communicate realistic timelines and data limits early rather than delivering an unreliable answer on time.
- Tested as
- Expectation-management scenarios.
- Trap
- Choosing the option that meets the deadline with untested data.
Side by side
Concepts candidates mix up, compared on the point that decides the answer.
Report vs. dashboard
| Concept | Meaning | Key difference | Common trap |
|---|---|---|---|
| Report | Static collection of data at a point in time | Quick to build and easy to share; does not update | Using it when stakeholders need to monitor changing data |
| Dashboard | Interactive, regularly refreshed view of live data | Tracks change over time in one place; needs upkeep | Building one for a one-off snapshot request |
Problem types that get confused
| Concept | Meaning | Key difference | Common trap |
|---|---|---|---|
| Making predictions | Using data to forecast a future outcome | Looks forward | Applying it to a question about past trends |
| Finding patterns | Using historical data to understand what happened | Looks backward at trends over time | Calling it a prediction |
| Discovering connections | Finding how separate entities or factors relate | Links different parties or variables, such as suppliers and delays | Calling it a pattern |
| Spotting something unusual | Identifying data that differs from the norm | Flags anomalies | Calling it categorizing |
Common spreadsheet errors
| Concept | Meaning | Key difference | Common trap |
|---|---|---|---|
| #DIV/0! | Division by zero or by an empty cell | The denominator is the problem | Blaming the numerator |
| #REF! | Reference to a cell that no longer exists | Caused by deleted rows or columns or a bad copy | Confusing it with #N/A |
| #N/A | Lookup value not found | The data is missing, not the reference | Rebuilding the formula when the key is simply absent |
| #NAME? | Function name or text not recognized | Usually a typo such as =SUMM() | Thinking it is a data-type error |
Exam traps
Each trap is an answer that looks right. Learn the clue that should change your mind.
-
Specific but closed
- What candidates pick
- 'Did sales increase in March?' as the best question.
- Why it looks right
- It is specific, measurable and time-bound.
- Why it's wrong
- A yes-or-no answer gives little to act on; a SMART question asks how much, why or for whom, so it supports a decision.
Clue to watch for An option that starts with 'did', 'is' or 'don't you think' usually invites a closed or leading answer.
-
Backward-looking prediction
- What candidates pick
- Making predictions, for a question about how sales changed over the last three years.
- Why it looks right
- Trend analysis feels like forecasting.
- Why it's wrong
- Examining historical change is finding patterns; making predictions uses data to estimate a future outcome.
Clue to watch for Past-tense time frames point to finding patterns; future decisions point to predictions.
-
Dashboard for everything
- What candidates pick
- Build a dashboard for a one-time year-end summary.
- Why it looks right
- Dashboards look more advanced and impressive.
- Why it's wrong
- A static report suits a fixed snapshot and is faster to produce; a dashboard earns its upkeep only when data changes and must be monitored.
Clue to watch for Words like 'one-time', 'end of quarter' or 'snapshot' versus 'monitor', 'track daily' or 'live'.
-
Fix the formula, not the data
- What candidates pick
- Rewrite a VLOOKUP that returns #N/A.
- Why it looks right
- An error looks like a formula problem.
- Why it's wrong
- #N/A means the lookup value was not found; the fix is often in the data, such as a missing key or extra spaces.
Clue to watch for The formula works for most rows and fails only for some values.
-
Silent scope change
- What candidates pick
- Quietly add the extra analysis a second stakeholder requested.
- Why it looks right
- It seems helpful and avoids conflict.
- Why it's wrong
- Unagreed changes cause scope creep and missed deadlines; the analyst should confirm priorities with the project owner and update the scope of work.
Clue to watch for A request that falls outside the agreed deliverables or timeline.
-
Deadline over accuracy
- What candidates pick
- Deliver preliminary numbers as final to meet an early deadline.
- Why it looks right
- It satisfies the stakeholder in the moment.
- Why it's wrong
- Wrong conclusions cost more than a short delay; the course favors explaining the trade-off and agreeing a realistic timeline.
Clue to watch for The data is incomplete or not yet verified when the stakeholder asks for results.
Answer framework
Apply these steps when a Course 2 question gives you a request, a stakeholder or a spreadsheet problem.
- Restate the business problem in one sentence and classify it as one of the six problem types. Note which element fails so you can explain the rewrite.
- Test the candidate question against each SMART element, rejecting leading, closed and vague wording. Matching the format to the need avoids wasted build time.
- Decide whether stakeholders need a static snapshot (report) or ongoing monitoring (dashboard), and name the metric.
- For spreadsheet items, read the error code or reference style first, then check the data before rewriting the formula.
- For people scenarios, clarify the goal, communicate limits and trade-offs, and record agreed changes in the scope of work before acting. Keep stakeholders informed of progress and any risks.
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
Begin with Module 1 and drill two things until they are automatic: the SMART elements, and the six problem types with one example each. Write ten business requests from your own work or news stories and classify them; the confusing pairs are finding patterns versus making predictions, and categorizing versus identifying themes.
Give Module 3 the most time, because it is the longest module and the most hands-on. Build a small sheet, copy formulas with and without $ signs, and deliberately trigger each error code so you recognize them on sight. Draft a one-page scope of work for an imaginary project to learn what belongs in it.
The underestimated area is Module 4. Stakeholder items look like common sense but reward a specific behavior: clarify, communicate trade-offs and document agreed changes. Afterward, take a practice set mixing SMART rewrites, problem types and spreadsheet errors, then move to Course 3, where the prepare phase begins.
FAQ
What does the A in SMART mean in the Google Data Analytics Certificate?
Action-oriented. Google's SMART questions are specific, measurable, action-oriented, relevant and time-bound, so the answer should support a decision or change.
What are the six problem types?
Making predictions, categorizing things, spotting something unusual, identifying themes, discovering connections and finding patterns. Questions usually describe a business need and ask you to name the type.
How is Course 2 assessed?
Each of the four modules ends with a graded module challenge, and the course also includes knowledge checks, self-reflections and hands-on spreadsheet activities. There is no separate proctored exam.
Which module is hardest?
Spreadsheet magic is the longest and most practical. Learners who have not used spreadsheets often need extra time on cell references, functions and error codes.
When should I choose a report rather than a dashboard?
Choose a report for a fixed snapshot that stakeholders review at a point in time. Choose a dashboard when they need to monitor data that changes regularly in one interactive view.
How much study time should I plan?
Coursera lists about 15 hours of content. Our editorial planning estimate is 20 to 30 hours to allow for spreadsheet practice and review.
How does Course 2 connect to the capstone?
The capstone case study starts with the ask phase: defining the business task and the key stakeholders. The SMART questions and stakeholder skills from this course are what that section is graded on in practice.
Other Data & Analytics certifications
Where the same subject shows up in other exams. Cards marked "Same subject" open the matching domain.
Test your Domain 2 knowledge
You have covered SMART questions, the six problem types, reports versus dashboards, spreadsheet errors and stakeholder communication. Test whether you can apply them to short workplace scenarios.
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