Companion to the 6-hour course. Visual, memory-first, active-recall. Same company throughout: Northwind Commerce (~$110M online retailer). Each block draws on the same shape — WHAT · WHY · PICTURE · ANALOGY · EXAMPLE · NUMBERS · GOOD vs BAD · MISTAKE · MEMORY HOOK · EXERCISE · RECALL · PROJECT USE — condensed to what each concept actually needs.
BUSINESS GOAL → BUSINESS QUESTION → KPI → DATA → ANALYSIS →
INSIGHT → STORY → DECISION → ACTION → OUTCOME → NEW KPI
KPI = WHAT is happening · ANALYSIS = WHY · STORYTELLING = SO WHAT · RECOMMENDATION = NOW WHAT
Memory hook: WHAT · WHY · SO WHAT · NOW WHAT.
BUSINESS GOAL → KEY QUESTIONS → KPIs
├ TARGET · ACTUAL · TREND
↓ VARIANCE ← WHAT
↓ ANALYSIS ← WHY
↓ INSIGHT
↓ STORY ← SO WHAT
↓ DECISION ← NOW WHAT
↓ ACTION (owner + date)
↓ IMPACT → new KPI
“What decision will this information help someone make?” No decision → why are we measuring it? This kills vanity metrics on contact.
Storytelling: turning analysis into a narrative — what happened, why it matters, what to do. DATA → MEANING → MESSAGE → ACTION.
KPI design: the right measurable indicators of whether an objective is met. GOAL → MEASURE → TARGET → STATUS → ACTION.
VANITY: Website Visits → "so what decision?" → none
ACTIONABLE: Conversion Rate → improve page/pricing/funnel/campaign
A metric becomes valuable only when it connects to a decision. Pair vanity totals with a rate or a cohort, or drop them.
BUSINESS GOAL → STRATEGIC KPI → DRIVER KPI → OPERATIONAL KPI → ACTIVITIES
Strategy watched monthly; operations daily.
PROFIT = REVENUE − COST
REVENUE = PRICE × VOLUME ; VOLUME = ORDERS → CUSTOMERS
E-commerce identity: Revenue = Traffic × Conversion × AOV. Any revenue move lives in one of those three branches. Tree answers why it moved; scorecard answers how are we doing.
Revenue = ₹100M → "compared with what?" vs Target · LY · LM · Budget · Forecast · Benchmark · Peer
KPI = VALUE + TARGET + TIME + COMPARISON + TREND + THRESHOLD + OWNER + ACTION
Revenue ₹125M · Target ₹120M · +4.2% · YoY +8% · ↑ · Owner: Sales · Action: maintain
Name · Definition · Formula · Numerator · Denominator · Unit · Grain · Time Period ·
Data Source · Target · Benchmark · Owner · Frequency · Threshold · Direction · Action
🎯 The denominator is where KPIs quietly lie. State it explicitly every time.
“What % of customers return?” → Returning ÷ Eligible. Margin = profit ÷ revenue = a keep-rate (for every ₹100 of sales, how much we keep).
LEADING = warning (pipeline, engagement, trial activation, complaints)
LAGGING = result (revenue, profit, churn)
Leading steers, lagging scores. A deck of only lagging KPIs is a rear-view mirror.
Leads → Qualified → Conversions → Revenue. Orgs over-watch outcomes, ignore the drivers they can move.
Revenue ↑ but Margin ↓; Conversion ↑ but Refunds ↑; Calls/day ↑ but CSAT ↓. Measure the outcome AND its guardrails.
PRIMARY KPI + GUARDRAIL KPI (Conversion ↑ + Refund rate must not rise). People optimize the metric, not the business — pair every incentive KPI with a quality guardrail.
COMPANY → BUSINESS UNIT → DEPARTMENT → TEAM → INDIVIDUAL. Every lower KPI ladders up to a business outcome.
KPI → OWNER → TARGET → REVIEW FREQUENCY → ACTION. GREEN on track · AMBER attention · RED action. Business-defined, not arbitrary colors.
The single metric capturing core value (Northwind: monthly repeat-purchasing customers). Don’t copy another company’s — it encodes their value.
Absolute vs relative (state the base) · percentage point vs percent (20%→25% = +5 pts AND +25% relative) · rate vs count · average vs median (mean lies with an outlier). Pick the stat from the question.
DATA → OBSERVATION → PATTERN → INSIGHT → IMPLICATION → RECOMMENDATION → ACTION
DATA ≠ INSIGHT: “Sales fell 12%” (data) vs “the drop is concentrated in the West, where Product A fell 25%” (insight).
SO WHAT? forces meaning · WHY? forces cause. Chain: WHAT HAPPENED → WHY → WHAT CAUSED IT → SO WHAT → NOW WHAT.
HEADLINE · EVIDENCE · DRIVER · IMPACT · RECOMMENDATION · ACTION
"Revenue is 8% below target. Three straight months below plan. West = 70% of the gap.
Quarter target at risk. Investigate West churn + Product A availability. Regional intervention in 2 weeks."
BAD title: "Monthly Sales Analysis"
GOOD title: "Sales are 8% below target, driven by West-region weakness."
Start with the answer, then prove it, then detail.
CONTEXT → PROBLEM → EVIDENCE → DRIVERS → IMPACT → RECOMMENDATION → ACTION. Seven patterns: Performance · Trend · Driver · Comparison · Exception · Opportunity · Risk.
Trend→line · Compare→bar · Rank→horizontal bar · Part-of-whole→stacked (donut 2–3) · Distribution→histogram/box · Relationship→scatter · Actual vs Target→bullet · KPI→card · Geography→map · Detail→table. Ask “what relationship must I show?”
Remove non-meaning ink (3D, heavy gridlines, rainbow). gray=context, blue=normal, green=positive, red=risk, amber=attention; one accent; never color alone. Biggest emphasis = most important — “what should they notice first?”
HEADLINE / KPIs → TREND → DRIVERS | EXCEPTIONS → RECOMMENDATION / ACTION
30-second test (what’s happening & what matters?) and 5-minute test (drill to root cause?). A dashboard that stops at LOOK (never DECIDE→ACT) is just a report.
WHO? → CARE ABOUT? → DECISION? → INFO? → ACTION? CEO=impact/risk/decision · CFO=revenue/margin/variance · Sales=pipeline/conversion/quota · Ops=volume/SLA/efficiency · Analyst=detail/drivers/root cause.
Root cause: SYMPTOM → PATTERN → DRIVER → ROOT CAUSE. Correlation ≠ causation (check mechanism, order, confounders). Hypothesis-driven. Exception analysis (top/bottom/outliers/change). Pareto 80/20. Contribution analysis and the variance/waterfall bridge (break a gap into additive drivers) — the single most persuasive executive artifact.
NUMBER → CHANGE → SIGNIFICANCE → BUSINESS IMPACT (Churn 7% → +2 pts → largest rise in 12 months → ~₹5M at risk). Compress numbers into a sentence, then show the chart.
KPI canvas: goal · success means · primary KPI · driver KPIs · leading KPIs · guardrails · target · owner · review freq · red threshold · action. Story canvas: WHAT · WHY · SO WHAT · NOW WHAT · who acts · by when.
One-slide: HEADLINE → KPI/TREND/TARGET → WHY (2–3 drivers) → SO WHAT (impact) → NOW WHAT (rec + owner + date). Speaking template: “The headline is X. Driven by Y. Impact is Z. We recommend A because B. Next action is C, owned by D, by [date].”
Sales · Marketing · Customer Success · Finance · Operations · HR — see the course doc for the full lists.
Vanity · ambiguous · too many · no owner/target/comparison · wrong denominator · lagging-only · conflicting · gameable.
Story checklist (headline, audience, KPI defs, validated numbers, comparison, drivers, root cause, impact, recommendation, owner, timeline). KPI checklist (relevant, measurable, defined, unambiguous, target, owner, comparable, timely, actionable, guardrail).
1 goal 2 decision 3 which KPI 4 vs what 5 what changed 6 why 7 impact 8 cost of nothing 9 what do 10 how know it worked
Revenue growing, profit suspected deteriorating. Scorecard: Revenue 110M🟢, Profit 18M vs 22M🔴, Margin 16.4% vs 21%🔴, Discount 9% vs 6%🔴, Retention 92%🟢. Decompose profit → revenue +10% but profit −5% because discounts rose and mix shifted to low-margin SKUs → margin compressed ~4.6 pts. Headline: “Revenue grew 10% but profit fell 5% — growth was bought via discounting, not earned. Review discount policy and low-margin SKUs; protect margin, not just sales.” Present in 5 min.
Be ready for: “Why care? How do you know? What’s causing it? What if we do nothing? What exactly do you recommend? How much impact? Who owns it? When will we know it worked?” Simulate from every seat.
GOAL → QUESTION → KPI → TARGET → ACTUAL → VARIANCE → TREND → DRIVER → INSIGHT
→ SO WHAT → RECOMMENDATION → NOW WHAT → ACTION (OWNER + DATE) → OUTCOME
KPI=WHAT · ANALYSIS=WHY · STORY=SO WHAT · RECO=NOW WHAT
Ready when, handed an unfamiliar problem, you can run: understand goal → smart questions → define success → KPI framework → KPI tree → data → analyze → drivers → test root causes → correlation vs causation → insight → headline → so-what → recommend → owner → deadline → measure outcome — and survive the CEO gauntlet.