📖 Content & Storytelling · Lesson 2 of 8

Using Data & Statistics

A number by itself means nothing. Context, translation, and a source are what turn a statistic into evidence the audience actually remembers.

📊

Three Rules for Numbers That Land

Most presenters use data badly — either listing too many stats or quoting raw numbers with no frame. Three rules fix this every time.

  • Pick ONE stat per claim — Multiple numbers in a row cause cognitive overload — audiences stop processing after the second one
  • Translate to human scale — Convert to a ratio, per-person figure, or time frame: '1 in 3' lands harder than '34%'
  • Source it briefly — 'McKinsey 2024 found...' is enough — reading a URL or full citation breaks your delivery
Diagram comparing weak raw statistics to translated and sourced versions using an worked example
Each row shows the same data point: raw (weak) → translated to human scale → sourced and connected to the audience.
💡 Tip: After you write any statistic, ask: can I turn this into a ratio, a per-person number, or a time frame? If yes, do it.
🎯

Weak vs Strong — Worked Example

Topic: 'Why most professionals will never use AI effectively.' Here is how the same statistic sounds weak, then strong.

Weak delivery (raw number, no source, no frame): "A lot of studies show that most professionals don't really use AI in a meaningful way. The numbers are actually pretty bad when you look at them."
⚠️ Watch out: Vague claims without a number feel like opinion. A number without a source feels made-up. Both lose the audience.

Strong Delivery — One Stat, Translated, Sourced

Here is the same point rewritten using all three rules.

  • One stat only — '1 in 3' — no second number follows it
  • Translated — '1 in 3' instead of '34%' — ratio is felt, percentage is processed
  • Sourced briefly — 'McKinsey 2024' — two words, full credibility
  • Connected to the room — "Two people in this room" — abstract data becomes present and personal
Strong delivery (one stat, human scale, brief source, audience hook): "McKinsey's 2024 report found that only 1 in 3 professionals who use AI every day actually drive measurable results from it. Pause. That means two people in this room — if three of you use AI — are not in that group. Yet. The difference is not the tool. It is how they prompt it."

Key Takeaways

  • 1Use ONE statistic per claim — more than two in a row and audiences stop absorbing them
  • 2Translate percentages to ratios or per-person figures: '1 in 3' lands harder than '34%'
  • 3Source every stat in two words: 'McKinsey 2024' or 'Stanford HAI 2025' is enough
  • 4Connect the data to the room: 'that means two people here...' makes abstract data present
  • 5Never read numbers off a slide — pause, let the audience see it, then speak to what it means