Latest

The THINK-ANALYZE-DECIDE Loop

A Framework for Going Beyond "The Right Chart"
Why This Framework Matters
This course taught you how to choose the right chart. But visualization is the final step, not the first. Before you ever open Excel, Tableau, or Python, you need a framework for thinking about what you're actually trying to understand. This is that framework.
The Problem with Starting with Charts
Most people jump straight to "What chart should I make?" But if you cannot clearly articulate the question, no chart will save you. Good analysis starts with clear thinking, not with visualization.
The Five-Step Framework
1

THINK: Clarify the Real Question

Before touching data or visuals, pause and ask:
  • What decision, judgment, or action might this inform?
  • What do we think is happening right now?
  • Who is affected, and how might their experience differ?
  • What assumptions are we making that need to be tested?
If you cannot clearly articulate the question, no chart will save you.
2

ANALYZE: Explore the Situation from Multiple Angles

This is where visualization lives, but it is not limited to one view.
  • Compare: Compared to what? (time, groups, benchmarks)
  • Diagnose: What might explain what I'm seeing?
  • Segment: Who is this working for, and who is it not?
  • Look for patterns: trends, variability, outliers, clustering
  • Check context: what data is missing, excluded, or lagging?
One chart shows a signal. Multiple perspectives reveal meaning.
3

INTERPRET: Make Sense Before You Explain

Slow down before jumping to conclusions.
  • What are at least two plausible interpretations?
  • What alternative explanations exist?
  • Where might bias or incomplete data be influencing me?
  • What would someone else reasonably question about this?
Interpretation is a thinking skill, not a visual one.
4

DECIDE: Determine What This Enables You to Do

Not every analysis leads to a decision, but every good analysis clarifies something.
  • What can we now say with more confidence?
  • What remains uncertain?
  • What action, experiment, or next question does this suggest?
  • What data would strengthen this decision next time?
Insight is not certainty. It is reduced uncertainty.
5

COMMUNICATE: Choose the Right Form Last

Only now ask:
  • What is the simplest way to show or explain this?
  • Does a chart help, or would a table, sentence, or question be better?
  • What does the audience need to notice first?
Visualization is the final step, not the first.
The Wrong Way vs. The Right Way
Starting with the Chart
  1. Open your data tool
  2. Make a chart
  3. Try different chart types
  4. Pick the one that "looks right"
  5. Show it to someone
Result: Pretty charts, unclear insights
Starting with the Question
  1. Clarify what decision this informs
  2. Explore from multiple angles
  3. Interpret what you're seeing
  4. Decide what this enables
  5. Choose the simplest visualization
Result: Clear insights, purposeful visuals
The Loop in Action: A Quick Example
Scenario: Your manager says "Show me our customer satisfaction scores"
THINK: What decision might this inform? Are we deciding whether to invest in customer service? Are we diagnosing why retention is low? Who are we comparing against?
ANALYZE: Look at scores over time (trend), by customer segment (who's satisfied vs. not), against industry benchmarks (compared to what?), and by service touchpoint (what explains the variation?).
INTERPRET: Are we seeing improvement because we fixed a problem, or because unsatisfied customers left? What alternative explanations exist? What's missing from this data?
DECIDE: This suggests we should focus on Segment A, not Segment B. Or: We need more data before we can decide. Or: This confirms our hypothesis and we can move forward.
COMMUNICATE: Now choose: line chart for trends? Grouped bars for segment comparison? A simple sentence if the answer is clear?
Detailed Examples: The Framework in Action
The best way to understand this framework is to see it applied to real scenarios. Here are two complete walk-throughs showing how to use THINK-ANALYZE-DECIDE-INTERPRET-COMMUNICATE for different types of questions.
Example 1: Employee Turnover Analysis
The Request: "Our HR director asks you to analyze employee turnover and present findings next week."
STEP 1: THINK - Clarify the Real Question
Don't jump to "What chart shows turnover?" Instead, ask:

  • What decision might this inform? Are we deciding whether to increase retention budgets? Change hiring practices? Address management issues?
  • What do we think is happening? Is turnover higher than before? Higher than industry? Higher in certain departments?
  • Who is affected? Are we losing top performers or low performers? New hires or veterans?
  • What assumptions need testing? Are we assuming all turnover is bad? That it's caused by pay?
Refined Question: "Which employee segments show unexpected turnover patterns, and what factors correlate with their departure?"
STEP 2: ANALYZE - Explore from Multiple Angles
Now use multiple views, not just one chart:

  • Compare: Current turnover (15%) vs. last year (12%) vs. industry benchmark (18%)
  • Segment: Turnover by department, tenure, performance rating, manager
  • Diagnose: Exit interview themes, time-to-departure after specific events
  • Look for patterns: Spikes after annual reviews? Clustering in certain teams?
  • Check context: Are we missing data on voluntary vs. involuntary departures?
What you discover: Turnover is up 3% overall, but 90% concentrated in employees with 6-18 months tenure in the Sales department under two specific managers.
STEP 3: INTERPRET - Make Sense Before You Explain
Slow down. Consider alternative interpretations:

  • Interpretation A: New Sales hires are poorly onboarded and leave when they realize expectations don't match reality
  • Interpretation B: Two managers have problematic leadership styles that drive attrition
  • Interpretation C: Our Sales compensation structure becomes uncompetitive after the first year
  • Where might bias influence me? Am I assuming management is the problem because I've heard complaints?
  • What would someone question? "Did these employees go to competitors or leave the industry entirely?"
Honest interpretation: "The pattern suggests either management or role clarity issues, but we need exit interview data and competitor compensation benchmarks to distinguish between explanations."
STEP 4: DECIDE - Determine What This Enables
What can we now do with more confidence?

  • What we know: The problem is concentrated in Sales, 6-18 month tenure, under two managers
  • What remains uncertain: Root cause (management, comp, role mismatch, or multiple factors)
  • Actions this suggests:
    1. Conduct focused interviews with recent Sales departures
    2. Compare these two managers' teams to other Sales managers
    3. Benchmark Sales compensation at 12-month mark
  • Data needed next: Exit interview transcripts, compensation comparisons, new hire survey results
Decision: "We have enough to focus retention efforts on Sales onboarding and management practices, but need more data before changing compensation."
STEP 5: COMMUNICATE - Choose the Right Form Last
Now decide how to communicate this:

  • Headline: "Turnover concentrated in early-tenure Sales employees"
  • Chart 1: Heatmap showing turnover by department and tenure (reveals Sales 6-18 month spike)
  • Chart 2: Bar chart comparing manager teams (shows two outlier managers)
  • Could skip: Overall turnover trend (doesn't add value given what we learned)
  • Better as text: "Recommended next steps" as bullet list, not a visual
Communication strategy: Lead with the insight ("Sales retention problem, not company-wide"), show focused evidence (two charts), recommend specific next steps. Don't bury the finding in generic turnover dashboards.
Example 2: Website Conversion Analysis
The Request: "Your marketing manager says: 'Our website conversion rate dropped last month. Figure out why.'"
STEP 1: THINK - Clarify the Real Question
Resist the urge to plot conversion over time immediately. Ask:

  • What decision needs this? Should we roll back website changes? Increase marketing spend? Fix technical issues?
  • What might be happening? Site redesign? Traffic quality shift? Seasonal pattern? Technical bug?
  • Who is affected differently? All visitors or specific segments? All pages or certain flows?
  • What do we assume? That conversion rate is the right metric? That one month is meaningful?
Refined Question: "What changed in visitor behavior or site experience that caused the conversion decline, and is it affecting our target customers?"
STEP 2: ANALYZE - Explore from Multiple Angles
Examine the situation from multiple perspectives:

  • Compare: This month vs. last month vs. same month last year vs. 6-month average
  • Segment: Conversion by traffic source, device type, new vs. returning, landing page
  • Funnel analysis: Where in the conversion path are people dropping off?
  • Look for changes: Site updates on specific dates? Marketing campaign changes? Page load times?
  • Check context: Did we change how we track conversions? Is traffic quality different?
What you discover: Overall conversion dropped from 3.2% to 2.7%, but mobile conversion dropped from 2.8% to 1.9% while desktop stayed flat. The drop started exactly when a new checkout page went live.
STEP 3: INTERPRET - Make Sense Before You Explain
Consider alternative explanations before concluding:

  • Interpretation A: The new checkout page has usability issues on mobile devices
  • Interpretation B: Mobile traffic quality decreased due to a marketing campaign change
  • Interpretation C: Mobile users are price-shopping more and converting on desktop later
  • Bias check: Am I assuming the site change caused this because I opposed it?
  • What's questionable? "Is the timing coincidence? Could there be a tracking issue?"
Honest interpretation: "Strong circumstantial evidence points to mobile checkout issues, but we should verify there's no tracking error and check for mobile-specific bugs before reverting the redesign."
STEP 4: DECIDE - Determine What This Enables
What actions does this support?

  • High confidence: Mobile checkout experience degraded starting with new page launch
  • Still uncertain: Specific usability issue causing the drop
  • Immediate actions:
    1. User testing on mobile checkout flow (within 48 hours)
    2. Verify analytics tracking is working correctly
    3. Review mobile checkout error logs and abandonment points
  • Contingency: Prepare to roll back checkout redesign if usability issues confirmed
Decision: "Investigate mobile checkout immediately. Don't change marketing or roll back redesign until we confirm root cause through user testing."
STEP 5: COMMUNICATE - Choose the Right Form Last
How should you present this finding?

  • Headline: "Mobile conversion dropped 32% after checkout redesign launch"
  • Chart 1: Line chart with annotation showing conversion by device type before/after launch date
  • Chart 2: Funnel showing where mobile users drop off in new checkout vs. old
  • Skip: Overall traffic trends (not relevant to this question)
  • Better as table: Specific conversion rates by device and date
Communication strategy: Start with the specific finding (mobile, not overall), show the timeline evidence (before/after chart), present diagnostic data (funnel), recommend urgent next steps. This is a decision-forcing brief, not an exploratory report.
What These Examples Teach Us
Notice what didn't happen in either example: we didn't start with "what chart should I make?" We started with clarifying questions, explored systematically, considered alternative explanations, determined what decisions this supported, and only then chose visualizations. The charts were servants of the thinking, not substitutes for it.

This is the discipline that separates analysts who make pretty dashboards from those who drive real decisions.
One-Line Takeaway for Your Desk
Clarify the question → Explore from multiple angles → Interpret carefully → Decide what this enables → Visualize last
Why This Matters in Your Career
The Skill That Sets You Apart
Anyone can learn to make a chart. What organizations value is the ability to think clearly about messy problems. This framework gives you a systematic way to:
  • Ask better questions before anyone asks you for an answer
  • Explore thoroughly instead of accepting the first pattern you see
  • Communicate with confidence because you've done the thinking
  • Build credibility as someone who delivers insight, not just charts
The people who get promoted are not the ones who make the prettiest dashboards. They are the ones who help their organizations make better decisions. This framework teaches you how.
Take This With You
Print this framework. Pin it to your wall. Reference it before your next analysis. The chart types you learned this week are tools. This framework is how you decide which tool to use, when, and why. That's the difference between doing data visualization and doing data-informed thinking.

Great! You’ve successfully signed up.

Welcome back! You've successfully signed in.

You've successfully subscribed to Turning Data Into Wisdom.

Your link has expired.

Success! Check your email for magic link to sign-in.

Success! Your billing info has been updated.

Your billing was not updated.