Defining the Question

Why this matters

Clear questions lead to clear insights. In Data Storytelling, the way you frame the question decides what data you collect, how you analyze it, and what decision you can confidently recommend.

• Real task: Translate a stakeholder’s vague ask (e.g., “Improve retention”) into a decision-ready question you can test.
• Real task: Document scope, metrics, and timeframe so your analysis stays focused and reusable.
• Real task: Align the question with an upcoming decision, not just a curiosity dive.

Concept explained simply

Defining the question means turning a broad goal into a precise, decision-ready statement. It clarifies who needs to decide, what success looks like, how you will measure it, and by when.

Mental model: DACOMT

• Decision: What decision will this answer enable?
• Audience: Who will use the answer to decide?
• Context: What business situation or hypothesis are we exploring?
• Outcome: What change or result do we want?
• Metric: How will we measure success?
• Timeframe: What period or deadline applies?

Template you can reuse: “To help [Audience] decide [Decision], we will assess [Context] to achieve [Outcome], measured by [Metric] over [Timeframe].”

Example of transforming a vague ask

Vague: “Why is churn high?”

Decision-ready: “To help the Product team decide whether to prioritize onboarding changes in Q2, we will analyze first-week activation behaviors to explain differences in 30-day churn rate among new users, measured as percent churn over the last 6 months.”

Worked examples (3+)

Example 1: Marketing campaign performance

Initial ask: “Did our campaign work?”

• Decision: Increase, pause, or shift budget.
• Audience: Marketing lead.
• Context: Paid social campaign for new signups.
• Outcome: Allocate next month’s spend effectively.
• Metric: Cost per activated signup; incremental lift vs. control.
• Timeframe: Last 4 weeks.

Decision-ready question: “To help the Marketing lead allocate next month’s budget, did the paid social campaign improve activated signups (vs. control) at an acceptable cost per activated signup over the last 4 weeks?”

Example 2: Reducing churn in a subscription app

Initial ask: “How do we reduce churn?”

• Decision: Which retention lever to prioritize (pricing vs. onboarding).
• Audience: Product manager.
• Context: Drop in retention among new cohorts.
• Outcome: Choose one lever to ship in next sprint.
• Metric: 30-day churn rate; activation completion rate.
• Timeframe: Last 3 new user cohorts; decide by Friday.

Decision-ready question: “To help the PM choose a retention lever by Friday, does improving onboarding completion correlate more strongly with lower 30-day churn than a pricing discount among the last 3 new user cohorts?”

Example 3: Operations delays

Initial ask: “Operations is slow, what’s going on?”

• Decision: Add staff vs. optimize workflow.
• Audience: Operations director.
• Context: Warehouse order lead times increased.
• Outcome: Restore service level.
• Metric: Median order lead time; queue length at packing; pick errors.
• Timeframe: Past 8 weeks; next capacity planning meeting.

Decision-ready question: “To help the Operations director choose staffing vs. workflow changes at the next capacity meeting, which stage (picking, packing, or shipping) most contributes to the recent increase in median order lead time over the past 8 weeks?”

How to define a decision-ready question (step-by-step)

1. Clarify the decision. Ask: “What decision will this answer help you make?”
2. Identify the audience. Ask: “Who will act on this and when?”
3. Set the context. Ask: “What changed, and what do we suspect?”
4. Define success. Ask: “If this goes well, what will be different?”
5. Pick a metric. Ask: “What precise metric reflects progress?”
6. Timebox it. Ask: “What period do we analyze, and when is the decision due?”

Stakeholder mini-script you can use

• “What decision are you trying to make?”
• “Who needs the answer and what format do they prefer (1 slide, brief memo)?”
• “What changed that triggered this question?”
• “What outcome would be a win in the next 4–6 weeks?”
• “Which metric would you trust to gauge that?”
• “What timeframe should we analyze?”
• “Any constraints (budget, tools, data access)?”

Templates and quick checks

Fill-in template:

“To help [Audience] decide [Decision] by [Deadline], we will analyze [Context/Hypothesis] to achieve [Outcome], measured by [Metric] over [Timeframe].”

Checklist before you proceed:

• Decision is explicit and time-bound.
• Audience is a real person/role, not “everyone.”
• Metric has a clear definition and unit.
• Timeframe is specified (lookback window and decision date).
• Scope is focused; you can answer it with available data.

Exercises

These mirror the graded exercises below.

Exercise 1: Rewrite a vague ask (ex1)

Scenario: A sales leader says, “Our enterprise pipeline feels weak. What’s going on?” Using DACOMT, rewrite this into a decision-ready question.

Expected output: A single sentence that states the decision, audience, context, outcome, metric, and timeframe.

Hints

• Decision likely involves where to focus reps or marketing budget.
• Metric candidates: qualified opportunities created, win rate, sales cycle length.

Exercise 2: Fill the gaps (ex2)

Stakeholder notes: “We want fewer support tickets from new users. Maybe the tutorial is unclear. Next planning meeting is in 10 days.” Fill in missing parts to form a decision-ready question. Include a suitable metric and timeframe.

Expected output: A one-sentence question using the template.

Hints

• Audience: Support lead or PM.
• Metric: ticket rate per 100 new users; first-week tutorial completion.
• Timeframe: recent new-user cohorts (last 4–8 weeks).

Common mistakes and how to self-check

• Vague goals (“improve engagement”) without a decision. Fix: Name the decision and by when.
• No metric or a fuzzy metric. Fix: Define a specific, calculable metric and unit.
• Overbroad scope. Fix: Limit to 1–2 hypotheses or a specific funnel step.
• Ignoring audience. Fix: State who decides and tailor the question to their choice.
• No timeframe. Fix: Add both analysis window and decision date.

Self-check mini task

Take one of your current tasks. Can you complete the sentence “This will help [Audience] decide [Decision] by [Date] using [Metric] over [Timeframe]”? If not, refine.

Practical projects

• Backlog cleanup: Convert 10 vague analytics requests into decision-ready questions. Share before/after.
• Team playbook: Create a 1-page DACOMT template and get stakeholders to try it on their next request.
• Retrospective: For your last project, rewrite the question and compare to what you actually answered. Note gaps.

Who this is for

• Data Analysts and aspiring analysts who present insights to business stakeholders.
• PMs, Marketing Analysts, Operations Analysts who translate goals into analysis.
• Anyone preparing data stories or decision memos.

Prerequisites

• Basic understanding of business goals and KPIs.
• Comfort reading simple metrics and trends.
• Willingness to ask clarifying questions.

Learning path

• 1) Defining the Question (this lesson).
• 2) Choosing Metrics and Definitions.
• 3) Forming Hypotheses and Assumptions.
• 4) Data Selection and Scope Control.
• 5) Narrative Structure and Visuals.
• 6) Presenting Recommendations and Trade-offs.

Mini challenge

Challenge: A PM says, “Trials aren’t converting to paid enough.” Write a decision-ready question that would guide a change in onboarding or pricing next sprint.

Show a strong example

“To help the PM prioritize onboarding vs. pricing in the next sprint, among trial users over the last 6 weeks, does improving completion of the ‘first value’ action have a stronger association with trial-to-paid conversion rate than offering a 10% discount?”

Next steps

• Use the template on your current request and get stakeholder confirmation in writing.
• Complete the exercises below and take the quick test.
• Apply DACOMT to one past project and note what you would change.

About the Quick Test

The quick test is available to everyone for free. Only logged-in users get saved progress and completion tracking.