Nancy Duarte's StorytellingResonate · Sparkline · Big Idea · DataStory

Great speeches don't state facts; they generate tension. Duarte's finding, from Lincoln to Jobs: every speech that has ever moved an audience oscillates rhythmically between two poles — What Is (today's reality) and What Could Be (tomorrow's possibility). Each oscillation pulls listeners a step further from their comfort zone, until they find themselves standing exactly where you meant them to stand.

This is the design of psychological potential energy. Pure data leaves the listener flat; pure vision feels weightless. Place them side by side — "today is heavy ⇄ tomorrow could be bright" — and a gravitational pull forms inside the audience. They aren't persuaded; they lean. The same physics underlies Kotter's change model and every Problem–Solution pitch. Duarte's contribution is to turn the instinct into a repeatable shape.

• Public talks, pitch decks, all-hands, product launches, change comms
• Anywhere the goal is to mobilize, not merely to report
• Skip for: pure technical specs, dry financials, urgent incident reports
• Use with care when the audience's What Is is already more painful than the one you sketch — you'll sound out of touch

• Only What Could Be — sounds like a motivational poster, no ground
• Only What Is — a diagnosis report, not a call
• Asymmetric poles (vivid pain, vague vision, or the reverse)
• One oscillation — not enough force to move anyone
• Failing to place the audience inside the What Could Be — they remain spectators

Nancy Duarte, Resonate: Present Visual Stories that Transform Audiences (2010) · Nancy Duarte, Illuminate (four-stage change narrative: Dream / Leap / Fight / Climb)

Duarte annotated hundreds of great speeches — Jobs, King, Churchill — line by line, marking each sentence as What Is or What Could Be, then connected the dots. The curve she calls a sparkline. The shape was eerily consistent: open in What Is, end in a New Bliss, oscillate the whole way through, and rise overall like the swell of a symphony. Before you write your next talk, sketch its sparkline first.

A sparkline makes time visible. A speech is not a document; it is an experience that unfolds minute by minute, with attention, emotion, and trust shifting at every beat. The sparkline forces you to ask, for each segment: which pole is the audience on right now, and which pole am I pushing them toward? When Jobs introduced the iPhone, the first thirty seconds promised to change everything (What Could Be). The next two minutes dwelled on the pain of juggling an iPod, a phone, and an internet device (What Is). Then he revealed the iPhone (a higher What Could Be). A textbook rising sparkline.

• Any talk longer than ten minutes — keynotes, TED-style talks, all-hands
• Investor pitches, product launches, annual strategy
• Skip for: under-five-minute updates, technical reviews
• Use with care for urgent internal comms — too much shape can feel insincere

• The whole talk glued to What Is — a diagnostic deck, not a speech
• The whole talk glued to What Could Be — castles in the air
• No New Bliss — the talk stops without landing
• No clear call to adventure — the audience never knows when to act
• Uneven cadence — dense oscillation early, flat finish

Nancy Duarte, Resonate, ch. 5 "Create Meaningful Content" · Steve Jobs, 2007 Macworld iPhone Keynote (Duarte annotates its sparkline minute by minute)

Before the sparkline, before the slides, before the script — write your Big Idea. Duarte's formula: Big Idea = your unique point of view + what's at stake for the audience. One sentence, no more than two lines. If you can't fit it, you haven't thought it through. Go think. This is the first gate in every Duarte workshop.

A unique point of view keeps the Big Idea from collapsing into platitude. "We should care about customers" is not a POV; "We should treat every return as a second chance to sell" is. Stakes keep it from collapsing into wish. "We should invest in AI" has no stakes; "If we don't rewrite the core in eighteen months, we lose the developer market" has. Multiply the two and you get resonance — the same logic behind the Heaths' Simple, McKinsey's "So what?", and a commander's intent.

• Before drafting any talk, article, memo, or deck — the "framing" step
• Aligning a team — only when everyone can recite the same Big Idea are you actually aligned
• Skip for: open-ended exploration — too early to lock in a conclusion
• Use with care if the sentence keeps changing — that means the thinking isn't done

• Mistaking a topic ("about customer success") for a Big Idea — that's a noun, not a claim
• No unique POV — true, but anyone could have said it
• No stakes — "I recommend X" with no "otherwise Y"
• Too long — over two lines because the thinking isn't done
• Jargon dressing up commonplace ("end-to-end value closure via downstream enablement")

Nancy Duarte, Resonate, ch. 4 "Define Your Audience" and "Develop Your Premise" · Chip & Dan Heath, Made to Stick, "Simple" chapter on Commander's Intent

Most data presentations fail not because the data is bad, but because the data is piled, not told. In DataStory, Duarte gives a three-act structure: Setup (establish the baseline), Conflict (reveal the turn), Resolution (point to a decision). Every chart, every number, has to earn its role in that story.

Numbers do not speak; the narrator gives them voice. A retention figure of sixty-seven percent can tell three different stories. (a) We beat the industry average, so spend more on acquisition. (b) We're behind the leaders, so rewrite onboarding. (c) We climbed from fifty to sixty-seven in three months, validating the new strategy — double down. Duarte's three acts force you to choose, explicitly, which decision this data is meant to serve — and only then to pick the chart, the comparison, and the path.

• Quarterly reviews, product data reviews, A/B readouts, board decks
• Anywhere you translate a dashboard into a decision narrative
• Skip for: raw data deliverables and research dataset releases — neutrality is the point
• Use with care when the data itself is highly uncertain — forcing a story becomes misleading

• Listing every metric — "everything" means "nothing"
• No Conflict — a smooth-rising story is forgettable
• Letting the chart speak while the presenter reads it — the audience doesn't know where to look
• No Resolution — the audience leaves without a next step
• Pretty-but-deceptive charts (truncated axes, bad scales, cherry-picked windows)

Nancy Duarte, DataStory: Explain Data and Inspire Action Through Story (2019) · Cole Knaflic, Storytelling with Data (the chart-craft companion) · Edward Tufte, The Visual Display of Quantitative Information