# The Science of Storytelling & Persuasion — Design Rules for Wealth-Management Materials

*Research memo. Sources are academic papers, books, and practitioner literature reviewed in Oct 2026 (web-verified). Use as the evidence base for the RM pitch deck, client-track deck, and monthly investment letters.*

---

## 1. The Evidence Base

### 1.1 Paul Zak — neuroeconomics of narrative (Claremont, Center for Neuroeconomics Studies)

**Source:** Zak, "Why Inspiring Stories Make Us React: The Neuroscience of Narrative," *Cerebrum*, 2015 (PMC4445577); Barraza & Zak, *Ann NY Acad Sci* 2009; Lin et al., *PLoS ONE* 2013; Barraza et al., *Biological Psychology* 2015.

Key findings:
- Narratives with a **dramatic arc** (tension → climax → resolution) cause measurable release of **cortisol/ACTH** (attention/arousal) and **oxytocin** (empathy/trust). A "flat" narrative with the same characters and no arc produced **no oxytocin, no cortisol, no empathy** — same characters, same footage, zero effect. **The arc, not the content, carries the chemistry.**
- Change in ACTH correlated with **attention paid**; change in oxytocin correlated with **concern for characters**. When **both** biomarkers rose, donations were **261% higher** than when one or both did not. Attention is necessary but not sufficient — it must be followed by emotional resonance.
- A physiological model (heart rate + skin conductance + vagal ECG) predicted **donation behavior with 82% accuracy** — story impact is physiologically measurable and predictable.
- Attention **waxes and wanes** even in a 100-second story; the brain scans away and refocuses. Peak attention occurs at the **climax/reveal**. Oxytocin (empathy) lags attention by ~30 seconds — audiences need ~30s of character exposure before caring.
- People **cannot articulate** what engages them: USA Today Super Bowl ad ratings (5,000 voters, 2013 & 2014) showed **zero correlation** with the physiological engagement measure ("ZEST"). Ask audiences what they want and they will describe the wrong thing.
- Effects of a story **outlive the story**: post-narrative donations occurred even when participants knew the story was fictional and actors staged it.
- One week later, **accurate recall was predicted by narrative transportation** — the single best recall predictor.

### 1.2 Green & Brock — Narrative Transportation Theory

**Source:** Green & Brock, "The Role of Transportation in the Persuasiveness of Public Narratives," *Journal of Personality and Social Psychology*, 79(5), 701–721 (2000). Four experiments: N = 97, 69, 274, 258.

Key findings:
- Transportation = absorption into a story (imagery + affect + attentional focus). Transportation **increased story-consistent beliefs and favorable evaluations of protagonists** across all four studies.
- Highly transported readers **found fewer false notes in a story** than less-transported readers — transportation suppresses critical scrutiny/counterarguing. (Design implication: a transported audience fact-checks less; therefore transport must never be earned with weak claims.)
- Belief change from transported narratives was **generally unaffected by whether the story was labeled fact or fiction** — plausibility and flow matter more than documentary framing.
- Transportation was **manipulable**: reducing it (processing instructions) reduced belief change. Transportation is a mechanism you can design for — vivid imagery, emotional beats, sustained focus.

### 1.3 Jennifer Aaker (Stanford GSB) — story vs. statistics

**Source:** "Harnessing the Power of Stories," Stanford Women's Leadership Innovation Lab / Voice & Influence series; Aaker's referenced studies (Escalas 2004, 2007; Small, Loewenstein & Slovic 2007; Lerner, Small & Loewenstein 2004).

Key findings:
- **Stories are remembered up to 22× more than facts alone.**
- Story + data together move audiences **both intellectually and emotionally**; data alone persuades intellectually only, and story alone persuades emotionally without proof. The design rule is conjunction, not either/or.
- Four elements of successful stories: **goal** (define what the audience should think/feel/do), **grab attention** (unexpected truth as hook), **engage** (protagonist + something at stake + arc), **enable** (memorable, retellable — your idea should be spreadable by the audience).
- Stories change **how others see you** — the stories told *about* the storyteller drive hire/buy/trust decisions. For an RM: the prospect retells your story to their spouse or family office.
- **Identifiable victim effect** (Small, Loewenstein & Slovic 2007): giving concentrates on one identified person; deliberation about statistics does not restore giving to statistical victims — deliberation *reduces* it. One concrete person beats an abstract population.

### 1.4 Chip & Dan Heath — Made to Stick (SUCCESs)

**Source:** Heath & Heath, *Made to Stick* (2007); heathbrothers.com (intro + Chapter 1 PDF). Evidence base includes empirical psychology (Tversky & Kahneman availability, noodle-kitchen urban legends as naturally "sticky" messages), the *Journal of Consumer Research* literature, and the popcorn case (CSPI: medium popcorn ≈ 37g saturated fat ≈ "a day and a half of artery-clogging fat"; the CBS *Eye on America* story it generated).

Key findings:
- Sticky ideas are **Simple, Unexpected, Concrete, Credible, Emotional, Stories** (SUCCESs). Non-sticky ideas typically fail on **Simple** (no single core message) or **Concrete** (abstract generalities).
- The **Curse of Knowledge**: once you know something you cannot un-know it, so experts systematically communicate abstractions the audience cannot decode. The single biggest failure mode of asset-management decks — the author is a PM writing for PMs.
- **Unexpectedness** opens attention gaps: break a schema (popcorn = junk food → "contains more fat than a bacon-egg-and-cheese plus sausage plus two orders of hash browns"), then close the gap.
- **Concreteness** beats abstraction: "a day and a half of saturated fat" lands; "37 g of saturated fat per 100 g" does not. Numbers must be translated into human-scale, sensory comparisons.
- **Credibility** comes from testable, specific details, not credentials alone ("Where's the beef?").
- **Data are not ideas.** The world's stickiest message ("the world is at war" / "movie popcorn is a fat bomb") is a *proposition the data support*, not the data themselves. Rule: extract the idea, let one number carry it, delete the rest.

### 1.5 Cognitive Load Theory (Sweller) → presentation design

**Source:** Sweller (1988, *Cognitive Science*); Sweller, van Merriënboer & Paas (1998); Mayer & Moreno (2002); Mayer, Heiser & Lonn (2001); Cambridge Handbook of Multimedia Learning (Mayer, ch. 12 & 15).

Key findings:
- **Working memory is severely limited**: Miller (1956) 7 ± 2 items; Cowan (2001) — the truer modern estimate is **~4 chunks**. Working memory is the bottleneck of every slide and every meeting.
- **Extraneous load kills learning**: Mayer's experiments showed that presenting *more* material (animation + narration + on-screen text) results in **less understanding** — redundancy effect. Reading slides aloud to the audience measurably degrades comprehension (Mayer, Heiser & Lonn 2001).
- Principles directly applicable to decks: **coherence** (delete interesting-but-irrelevant material), **signaling** (visually mark what matters), **redundancy** (never narration + identical text), **spatial contiguity** (label next to the thing it labels, not in a legend), **temporal contiguity** (explain a number when it is on screen, not before or after).
- Split-attention: any layout forcing the audience to integrate two physically separated sources (chart left, footnotes right, legend on top) burns working memory on integration instead of insight.

### 1.6 Attention-span evidence

**Source:** Medina, *Brain Rules* ("10-minute rule": audiences disengage from medium-interest lectures roughly every 10 minutes; the claim is contested — Bunce et al. 2010 found attention fluctuates rather than decays monotonically, with drop-offs in the first ~30 s and mid-lecture — but every study agrees attention is episodic, not sustained).

Design consequence: **re-hook every ~10 minutes** (or per ~4–5 slides): a question, a number, a story pivot, a chart reveal. Attention is restarted by novelty, not by continuing louder.

### 1.7 Storytelling with Data (Cole Nussbaumer Knaflic)

**Source:** Knaflic, *Storytelling with Data* (Wiley, 2015); storytellingwithdata.com blog ("my guiding principles," 2017; "the 3-minute story"; "what's the Big Idea?").

Key findings:
- Seven guiding principles: (1) be clear on **intent** (exploratory vs. explanatory — client-facing work is always explanatory: have a point of view and make it clear); (2) the right graph creates an **"aha" moment** — looking at it should not feel like work; (3) **don't overcomplicate**; (4) **get rid of the non-essential** ("perfection is when there is nothing left to take away"); (5) **make it clear where to look** (visual hierarchy: emphasize the one element that matters, grey out the rest); (6) **words make a graph accessible** — every chart gets a title that states the takeaway and a footnote with source/methodology; (7) **audience trumps all else**.
- The **3-minute story** and **Big Idea**: if you can say it in three minutes without slides and it fits in one sentence (point of view + what's at stake), the deck is built on a spine. If you can't, you have a data dump.
- "Data in a spreadsheet or facts on a slide aren't things that naturally stick with us — they are easily forgotten. Stories are" (ch. 6, *tell a story*).

### 1.8 Application in financial services specifically

**Sources:** Susan Weiner CFA (investmentwriting.com quarterly-letter survey, practitioner interviews); InvestmentNews ("Tell a story, keep a client"; "Why Storytelling Matters in Advisor-Client Relationships"); Janus Henderson practice management ("Why storytelling is an advisor's most powerful communication tool"); State Street Global Advisors ("The power of storytelling for financial advisors"); Kitces.com on client newsletters; Guy Kawasaki 10/20/30 rule.

Key findings:
- Weiner's survey of investment-letter readers: the one-words chosen most were **"clarity," "insight," "candor."** Clients value **what went right AND what went wrong** — candor is itself a trust signal (the Buffett precedent).
- Ranked importance of letter components: (1) manager's strategy, (2) past-quarter portfolio performance review, (3) outlook, (4) graphs/tables, (5) client-specific returns, (6) security comments, (7) sector strategy, (8) market recap. **The market recap is near the bottom** — readers can get it from Bloomberg; they want what they *cannot* get elsewhere: your attribution and your judgment.
- Length: **>40% of respondents want ≤2 pages**; 5+ pages is the least popular. "Investors want you to tell them what THEY need to know, not everything YOU know."
- Kawasaki 10/20/30: **10 slides, 20 minutes, no font < 30 pt**. "A normal human being cannot comprehend more than ten concepts in a meeting." Font rule forces salience: if it doesn't fit at 30 pt, it isn't a point.
- Industry consensus (Janus Henderson, SSGA, InvestmentNews): stories make abstract financial concepts (diversification, drawdown, sequence risk) **felt**, build trust faster than numbers, and differentiate firms whose products are otherwise commoditized. But all of them pair story with data — story alone reads as salesmanship; the compliant, fiduciary context demands the data be *there*, subordinated to the narrative.

---

## 2. Translated Design Rules

### 2.1 Pitch deck structure (RM-track and client-track)

| Rule | Basis |
|---|---|
| **10 slides, hard cap (12 for the client track with appendix)** | Kawasaki 10/20/30; Zak (attention is episodic); Sweller (4-chunk working memory — 10 slides ≈ one concept per slide is the ceiling) |
| **One core message ("Big Idea") before any slide is built**; it fits in one sentence and survives the 3-minute no-slides test | Knaflic Big Idea / 3-minute story |
| **One idea per slide; ≤ 30 pt font; ≤ ~6 lines of text max** | Kawasaki; Mayer redundancy principle |
| **Never read the slide.** Slide = headline takeaway + one visual; voice = the story | Mayer redundancy effect (reading identical text aloud degrades understanding) |
| **Re-hook every ~10 minutes** — open slides 1, 5, 9 with a question, surprise number, or story pivot | Medina 10-minute rule; Zak (attention waxes/wanes) |
| **~30–40 seconds of relationship/character setup before the first big claim** — trust (oxytocin) lags attention by ~30 s | Zak timing findings |
| **Climax placement: the strongest proof point (e.g., the 2008 drawdown experience) lands as a dramatic reveal in the last third, not buried on slide 3** | Zak peak-attention-at-climax |
| **Emphasize ONE element per chart; grey everything else; title states the takeaway as a sentence** | Knaflic; Mayer signaling |
| **Test the deck on a non-expert** and ask what they remember — people can't self-report what engages them, so test recall, not opinion | Zak (USA Today zero-correlation finding); Aaker "test your stories" |

### 2.2 Number presentation — what to show, what to cut

**Show (each translated to human scale):**
- The **one headline number** per section, stated as an idea: not "annualized return 9.4% since inception" but "every franc our clients invested has grown ~2.4× over the period the S&P did 2.0× — with roughly half the worst-year drawdown." (Heath concreteness; Aaker story+data.)
- **The drawdown story, not the drawdown table**: 2008 as a narrative ("we cut exposure in month X; clients who left the market in fear locked in the loss — here is what the return gap looked like five years later"). One identifiable client anecdote beats a 12-column attribution table (Small/Loewenstein/Slovic identifiable-victim effect; Aaker 22× recall).
- **Cost translated to lifestyle**: "1.1% all-in fee on CHF 2m costs CHF 22,000/yr — about a family ski holiday, every year, forever." Concrete, unexpected, testable (Heath SUCCESs: Concrete + Unexpected + Credible).
- **Attribution in plain language**: "we made money because…, we lost money because…" — the single most requested item in Weiner's survey ("what went right, what went wrong").
- Per chart: **one series highlighted, one takeaway title, source footnote** (Knaflic principles 5–6).

**Cut:**
- Multi-line return tables, 3-year vs 5-year vs 10-year grids, benchmark-plus-index-plus-peers comparisons — relegate to appendix or one page in the follow-up pack (extraneous load; coherence principle; Mayer's "more material = less understanding").
- Any number you cannot attach a meaning to ("we hold 47 positions across 14 sectors") — data without an idea is noise (Heath: data are not ideas).
- Jargon + acronym + decimal precision ("-1.37% ex-CHF EM USD-hedged") — Curse of Knowledge; write at the client's level, not to impress peers (Weiner survey respondents, verbatim).
- Anything that forces split-attention: no legend-on-top + footnotes-right + cross-referenced notes. Label directly on the chart (Mayer spatial contiguity).

### 2.3 Emotion/caution balance in a compliance- and fiduciary-driven context

The wealth-management version of the emotional beat is **not** sentiment — it is **stewardship under stress**. Design rules:

- **Emotional vehicle = client outcomes under a real drawdown** (the identifiable person: "a client in 2020" anonymized per compliance), not fear-marketing or greed-marketing. Zak's finding cuts both ways: the arc triggers empathy for *characters*; your protagonist must be the client, never the firm. This is both the highest-leverage design choice and the compliance-safe one.
- **Candor is a persuasion device, not a risk**: the single most-cited trust words were clarity/insight/**candor**; "what went wrong" sections outperform perfection (Weiner survey; Green & Brock — transported audiences detect fewer false notes, but in a fiduciary setting an eventually-discovered omission destroys trust retroactively; candor is the asymmetric bet). Rule: every performance claim is paired with a named risk or a period where the strategy underperformed.
- **Never let transportation do the fact-checking's job.** Because transported audiences counterargue less (Green & Brock Exp. 2), all forward-looking claims must carry method and source footnotes *in the deck itself* — you are legally liable for claims the audience didn't scrutinize.
- **Fear is allowed only as challenge, then immediately answered**: state the risk plainly (sequence risk, currency, concentration), show the mechanism that manages it, end on control. Open loops without resolution raise cortisol with no oxytocin — anxiety, not trust (Zak: both biomarkers must rise).
- Regulatory superlative/benchmark rules apply regardless of narrative framing — compliance reviews the *claims*, the narrative design never widens what may be claimed.

### 2.4 Monthly/quarterly investment letters — structure people actually read

From Weiner's survey + Knaflic + Heath + Green & Brock:

1. **Length: ≤ 2 pages body**, charts allowed beyond if truly one-message-each. (>40% of readers want ≤2 pages; 5+ pages least popular.)
2. **Open with the Big Idea in 1–3 sentences** — the quarter's one thought, stated as a point of view, before any market recap. (3-minute story; readers "pull out the important nuggets quickly.")
3. **Then "what we did and why"** — attribution in plain language: what went right, what went wrong, what we changed. This is the #1 and #2 ranked content (strategy + performance review); the #8-ranked item (market recap) gets two sentences max or a sidebar.
4. **One table maximum** (performance + positions), everything else prose + at most two takeaway-titled charts.
5. **One concrete client-outcome story per letter** — a holding thesis told as a narrative ("why we bought it, what happened, what we did"). Stories are 22× more memorable (Aaker); a letter with zero stories is a letter that is forgotten.
6. **One forward-looking risk stated candidly**, with what would change our mind. (Candor; falsifiable credibility — Heath's "testable credential.")
7. **Write at the client's level, in your own voice**; no words that fly over their head (Weiner respondents verbatim). A busy reader must understand it in one pass.
8. **Vary the format, not the discipline**: the skeleton is fixed, the content rotates. Predictable structure is what makes a monthly letter *read* — the reader learns where their answer lives.

---

## 3. The Narrative Arc Template

Derived from Zak's dramatic-arc neurochemistry (attention ↑ → empathy ↑ → action), Green & Brock's transportation, and Aaker's four elements (goal / hook / protagonist-at-stake-arc / retellability).

**The five-beat arc:**

1. **HOMELAND — where the client stands.** Concrete, in their terms. (Hook: an unexpected truth about their situation.)
2. **TENSION — the threat or gap they haven't fully seen.** Raises attention (cortisol/ACTH). Must be real, not manufactured fear.
3. **THE TURN — the moment a decision had to be made** (a market event, a life event, a past failure). This is the story's climax — place the strongest evidence here.
4. **THE PATH — your method as the resolution mechanism.** Data live here, one number per beat, subordinated.
5. **THE NEW NORMAL + CALL — life after, and the specific next step.** Ends the open loop; gives the retellable sentence ("the one thing they say about you").

**Fill-in skeleton — RM-track (pitch to a prospective RM joining the firm):**

1. **Homeland:** "You've spent ten years building a book of HNW clients in Zurich. The firm you're at charges 1.2%, gives you four systems, none of which talk to each other."
2. **Tension:** "Your clients' next decade is decided by whether their advisor can show drawdown discipline — and your current platform makes that manual, slow, and invisible to clients."
3. **The Turn:** [reveal] "In March 2020 our whole client book got a one-page portfolio review within 72 hours — risk overlays had rebalanced automatically, because the system, not the person, holds the discipline." *(climax = the strongest proof, revealed, not stated)*
4. **The Path:** "We combine fundamentals-driven research with a price-action overlay — one number per beat: median drawdown vs peers, client retention rate, time-to-report."
5. **New Normal + Call:** "Your clients would hear from you before they hear from the markets. The first step: two clients you'd migrate as a pilot, with our team running the letters."

**Fill-in skeleton — Client-track (RM presenting to an HNW prospect):**

1. **Homeland:** "You've built CHF 8m over thirty years. You could stop working tomorrow — but your real question is whether the money survives three more decades and two inheritances intact."
2. **Tension:** "The risk you haven't priced isn't return — it's sequence: a 30% drawdown in year one of your retirement changes everything, even if average returns are fine afterward." *(unexpected truth = hook)*
3. **The Turn:** [client story, anonymized] "A client retired in 2007. Six months later the market fell 40%. Because the glide-path had de-risked two years prior and the overlay cut equity exposure in the drawdown, she never sold a position in panic — she flew to South America that winter."
4. **The Path:** "Three mechanisms, one sentence each: diversification you can point to, a written de-risking calendar, and quarterly letters that tell you what we did and why — including when we were wrong."
5. **New Normal + Call:** "Your money is boring in the best sense: rules decide, you live. Next step: we review your current portfolio against these three tests — no commitment, one page back."

**Arc rules:** the client is the protagonist (Aaker: protagonist need not be the storyteller — best if it is the customer); something must be at stake (Zak empathy requires stakes); the reveal goes in beat 3, not beat 1 (Zak peak attention at climax); every arc ends with a closed loop and one concrete ask (Zak: transported audiences act *after* the story — the ask rides the post-narrative window).

---

## 4. Anti-Patterns and Why They Fail (per the literature)

| Anti-pattern | Why it fails (source) |
|---|---|
| **The data dump** (every chart, every year, every benchmark) | Extraneous load: Mayer showed adding material *reduces* understanding; Knaflic: "get rid of the non-essential"; Heath: data are not ideas — without a core message the audience extracts nothing; working memory holds ~4 chunks (Cowan), a 12-row table consumes it entirely before insight starts. |
| **The 40–60 slide deck** | Kawasaki: "a normal human cannot comprehend more than ten concepts in a meeting… if you need more than ten slides you don't have a message"; Zak: attention is episodic and unrecoverable — a deck longer than attention cycles guarantees dead sections; Weiner: readers won't start long, boring documents ("they simply won't bother"). |
| **Tables as content** | Tables are exploratory, client materials are explanatory (Knaflic intent rule); split-attention (chart + legend + footnotes in separate places) burns integration capacity (Sweller/Mayer); tables have no arc → no transportation → no belief change (Green & Brock) → forgotten within a week (Zak: recall predicted only by transportation). |
| **Reading slides aloud / text-wall slides** | Redundancy effect: narration + identical on-screen text measurably degrades comprehension (Mayer, Heiser & Lonn 2001); audience reads faster than the speaker speaks, desynchronizing the room (Kawasaki); font <30 pt = the message was never distilled. |
| **Numbers without translation** ("37 g saturated fat" vs "a day and a half of artery-clogging fat") | Curse of Knowledge (Heath): the author cannot feel the audience's ignorance; abstract statistics do not trigger empathy — deliberation over statistical victims *reduces* giving (Small/Loewenstein/Slovic); concrete, human-scale comparisons are the sticky form (Heath Concrete). |
| **Perfection-narrative ("we're always right")** | Readers rank candor among the top values and explicitly want "what went wrong" (Weiner survey); claimed infallibility is a false note that transported audiences would otherwise never check — and trust, once the note is found, is unrecoverable. |
| **Fear without resolution** | Cortisol without oxytocin: attention without empathy produces anxiety, not action (Zak — both biomarkers must rise for the 261% effect); in a fiduciary context open fear-loops read as sales pressure and fail compliance review. |
| **Asking the audience what they want, then building that** | Zak's Super Bowl studies: stated preference had zero correlation with measured engagement; build the arc, then test *recall* on real people, not stated preferences. |
| **Firm-as-protagonist stories** | Aaker: the protagonist is the customer; Zak: empathy attaches to characters with stakes. "Our platform, our team, our track record" has no character for the audience to transport into. |

---

## 5. Quantified Findings — Quick Reference

- **22×** — recall advantage of stories over facts alone (Aaker, Stanford GSB).
- **+261%** — donations when a story raised *both* arousal (ACTH) and empathy (oxytocin) vs. when one or neither rose (Zak lab).
- **+56%** — donations under oxytocin infusion vs placebo across 16 PSAs (Lin et al., *PLoS ONE* 2013).
- **82%** — accuracy of predicting donation behavior from peripheral neurologic response to a story (Zak lab).
- **0 correlation** — stated ad preference (USA Today, ~5,000 raters) vs physiologically measured engagement, replicated 2013 and 2014 (Zak lab).
- **~30 s** — lag of empathy (oxytocin) behind attention at story start (Zak lab).
- **~4 chunks** — working-memory capacity (Cowan 2001; vs Miller's 7±2, 1956).
- **10 minutes** — attention re-hook interval (Medina, *Brain Rules*; episodic per Bunce et al. 2010).
- **10 slides / 20 min / 30 pt** — pitch-deck constraint (Kawasaki).
- **≤ 2 pages** — preferred investment-letter length, >40% of readers; 5+ pages least popular (Weiner survey).
- **Letter content rank order** — strategy > performance review > outlook > charts > client returns > security comments > sector strategy > market recap (Weiner survey).
- **Transportation scale** — Green & Brock (2000), 4 experiments (N=97/69/274/258): transportation ↑ story-consistent beliefs, ↓ flaw detection, fact/fiction label largely irrelevant.

## 6. Source List

1. Zak, P.J. (2015). "Why Inspiring Stories Make Us React: The Neuroscience of Narrative." *Cerebrum* (Dana Foundation), PMC4445577.
2. Lin, P-Y. et al. (2013). "Oxytocin Increases the Influence of Public Service Advertisements." *PLoS ONE* 8(2).
3. Barraza, J.A. et al. (2015). "The Heart of the Story: Peripheral Physiology During Narrative Exposure Predicts Charitable Giving." *Biological Psychology*.
4. Kosfeld et al. (2005). "Oxytocin Increases Trust in Humans." *Nature* 435.
5. Green, M.C. & Brock, T.C. (2000). "The Role of Transportation in the Persuasiveness of Public Narratives." *JPSP* 79(5), 701–721.
6. Aaker, J. "Harnessing the Power of Stories." Stanford GSB / Women's Leadership Innovation Lab (+ referenced Escalas 2004/2007; Small, Loewenstein & Slovic 2007, *OBHDP* 102; Lerner et al. 2004, *Psych Science* 15).
7. Heath, C. & Heath, D. (2007). *Made to Stick: Why Some Ideas Survive and Others Die*. Random House.
8. Sweller, J. (1988). "Cognitive Load During Problem Solving." *Cognitive Science* 12; Sweller, van Merriënboer & Paas (1998). *Educational Psychology Review*.
9. Mayer, R. & Moreno, R. (2002). "Aids for computer-based multimedia learning." *BJET*; Mayer, Heiser & Lonn (2001). "Cognitive Constraints on Multimedia Learning." *JEP: Applied* 7(1).
10. Miller, G. (1956). "The Magical Number Seven." *Psych Review*; Cowan, N. (2001). "The Magical Number 4." *BBS* 24(1).
11. Medina, J. (2008). *Brain Rules* (attention chapter); Bunce et al. (2010). "Attention During Lectures Beyond Ten Minutes."
12. Knaflic, C.N. (2015). *Storytelling with Data*. Wiley; storytellingwithdata.com blog (2014–2017 posts).
13. Weiner, S. (CFA). "Ideal quarterly investment letters: Meaningful, specific, and short." investmentwriting.com (survey of investment-letter readers).
14. Kawasaki, G. "The 10/20/30 Rule of PowerPoint." guykawasaki.com.
15. Janus Henderson, SSGA, InvestmentNews, Kitces.com — advisor storytelling practice literature (2010–2024).