What it is
The default form: a 30–150 page slide deck addressed to fellow shareholders, the board, and proxy advisors. Designed to be printed, circulated, read cold, and debated by committee. Written to survive without a live presenter.
This is the centre of gravity of the adversarial-research canon. Pershing Square · Canadian Pacific, Elliott · Arconic, Starboard · Darden, Trian · Disney all use this form. When in doubt, choose this.
When to use it
- ✅ Campaign will culminate in a board vote or public capital event
- ✅ Your ask requires sustained argument (structural change, CEO replacement, large capital decision)
- ✅ Institutional holders are the primary audience
- ✅ You expect the target to publish a counter-deck — you need to occupy every frame first
- ❌ For a 2–10 page public communication, use
forms/letter-to-board.md - ❌ For a short-seller thesis, use
forms/short-report.md - ❌ For internal pre-decision circulation, use
forms/research-note.md
Length and pacing
- Short deck: 30–60 pages. Single-thesis campaigns, early-stage.
- Standard deck: 60–120 pages. Most proxy-contest decks sit here.
- Long deck: 120–300 pages. Full campaign compendium; fractal 8-block structure where each section is itself an 8-block mini-deck.
Page count discipline: every slide must survive the "what does this slide contribute" test. If you cannot answer in one sentence, cut it.
Architecture
The 8-block sequence in slides/slide-architecture.md applies directly:
1 Cover (1 page)
2 Thesis headline (1 page)
3 Executive summary (1–2 pages — the three reasons)
4 Situation (5–8 pages — the neutral frame)
5 Complication (15–30 pages — the case)
6 Question (1 page — the pivot)
7 Answer (10–15 pages — demands + valuation + precedent)
8 Closing ask (1 page)
+ Appendix (unbounded — methodology, sources, bios)
Long decks apply the same pattern inside each major section.
Reader behaviour to design for
The institutional reader:
- Flips to the cover (2 seconds)
- Reads the three reasons (15 seconds)
- Jumps to the valuation slide (1 minute)
- Jumps to the closing ask (15 seconds)
- If still interested, starts at slide 1 and reads in order
Your deck must work at each of these touchpoints. A deck that only works when read start-to-finish has failed 90% of its readers.
Design implications:
- The cover's three reasons must carry the full thesis. Tested: can a PM who only reads the cover recite your thesis correctly?
- The valuation slide must be standalone. Label axes, show method, state the bridge — assume the reader arrived via keyword search.
- The closing ask must name the verb, the actor, and the deadline.
Voice
- Adversarial when naming failures, analytical when presenting facts. Switching registers signals the author knows which mode they're in.
- Attack the decision, not the person. Villain-named directors and CEOs appear by name, but the argument is always about what they chose, not who they are.
- Let management's words do the work. Prior CEO quotes, investor-day
promises, and proxy-statement language are more credible than your
paraphrase of the same. Use
patterns/ceo-quote-contradiction.md. - No hedging in headlines, all hedging in footnotes. Headlines are claims; footnotes are where you concede edge cases and cite sources.
Distribution
The deck is published at a campaign domain (e.g.
restorethemagic.com, streamline66.com), filed on EDGAR as
DFAN14A, and simultaneously emailed to top-50 institutional holders.
ISS and Glass Lewis receive it with an accompanying cover memo.
Timing: anchor to a scheduled catalyst (Annual Meeting, earnings, Investor Day). Avoid holiday weeks. Avoid peer earnings days.
The appendix is not optional
Institutional readers will read the appendix. Short sellers and proxy advisors will stress-test every cited figure. The appendix contains:
- Peer-set definition and justification
- Full source citations with URLs and access dates
- Methodology notes for every derived number
- Biographies of proposed director nominees
- Regulatory filings referenced
- Counter-arguments considered and rejected (inoculation)
A deck without a rigorous appendix reads as a memo, not a campaign.
Anti-patterns
- Decorative stock photography. Top-tier decks never use generic imagery on the cover. Use the target's logo, a chart, or type alone.
- Dense text slides. If a slide has more than 60 words, split it.
- Too many KPIs per chart. One metric per chart. If you need two, make two charts.
- Closing ask as a thank-you. The final slide is a demand with a verb, not a closing pleasantry.
- Missing the pivot slide (Block 6). Without the Question slide, the Complication bleeds directly into the Answer and the reader never catches the hinge.
Exemplars
- Pershing Square · Canadian Pacific (Feb 2012) — 112 pages, the archetype
- Starboard · Darden (Sep 2014) — 294 pages, the long-deck template
- Elliott · Arconic (Apr 2017) — 336 pages, long-deck with visual craft excellence
- Trian · Disney "Restore the Magic" (Mar 2024) — 133-page white- paper version of the activist-deck form
- Elliott · Phillips 66 (Apr 2025) — modern proxy-era canonical deck; multi-document campaign
- Oasis · Kao Corporation (Feb 2025) — Japan-context activist deck
Full list: examples/masterclass.md
See also
slides/slide-architecture.md— the 8-block sequence in detailslides/cover-slide.md— the first pageslides/closing-ask-slide.md— the last pageslides/visual-craft.md— typography, density, colourcampaigns/initial-thesis.md— opener positioningcampaigns/proxy-fight.md— when the deck supports a vote