Version Compatibility

Reference framework authored 2026-06 from a synthesis of public sources. No code dependencies. Update when:

• The conference changes the poster dimensions (e.g., landscape to portrait to wide)
• The venue introduces a digital poster format and the scan-time assumption changes
• A peer-reviewed update to the Loui & Smith poster-specific guidelines is published

If dimensions, time, or font-size constraints differ from the defaults below, resize proportionally; do not change the grid structure.

Poster

A conference poster is a document designed for three reading modes with different times: the 2-minute scan (the passerby who glances at 20 posters and stops at 3), the 10-minute deep read (the collaborator who reads every figure), and the 3-minute Q&A (the presenter who must defend every element of the deciding experiment). A poster that does not serve all three modes serves none of them.

A poster is not a paper printed on a single page. It reorganizes a paper into the visual register, where the reader's eye follows a grid, not a narrative. The default failure mode is the single-column paper-flow; the alternative is the single-take-home-message grid.

The Three-Mode Design Test

Design the poster against three test scenarios.

A poster that serves the scan is a good poster. A poster that serves the deep read is useful for collaboration. A poster that serves Q&A is a defensible claim. All three are necessary.

The Grid Architecture

The poster uses a Decide-Support-Imply grid, derived from Loui & Smith's five-act talk structure but reorganized for the visual register. In a 48-inch x 36-inch landscape poster at A0 size, map the space as follows:

Total: ~600 words on a poster. The temptation to bring a 1500-word document is the temptation to bury the message; resist it. Every word on the poster should be legible from 1.5 meters (5 feet). If a reader must step within arm's reach to read a word, the word is not on the poster.

The 60-Second Visual Test

A poster that can be understood in 60 seconds is a poster. A poster that requires 10 minutes to understand is a paper. The 60-second test is the filter.

1. 15 seconds on the title bar + take-home. The reader sees the title and the result sentence. If this does not capture attention, the scan is over.
2. 30 seconds on the decide panel. The reader looks at the deciding figure. If the figure does not carry the result, the deep read does not begin.
3. 15 seconds on the imply bar. The reader sees the implication. If this does not justify the deep read, the scan ends.

If the 60-second test is too aggressive, the poster serves only the deep-read mode and the author spends the poster session alone. A well-designed poster has a reader rate of 10+ readers per hour in scan mode; a paper-as-poster has 2-3.

Panel Content Rules

Title Bar

• Title is 10-12 words, title-case, no jargon. "Epigenetic adaptation drives temozolomide resistance in glioblastoma" is acceptable; "METTL3-mediated m6A modification of MYC mRNA drives temozolomide resistance" would be the title, and "METTL3" is defined in the body.
• Authors: surname + initials, underlined presenter. Use a standard abbreviation for the institution.
• Affiliation: one line, numbered if needed. Do not repeat affiliations on every panel.

Take-Home Bar

One sentence that states the result, one sentence that states the implication. The bar should be readable from 3 meters (10 feet) — use a high-contrast background color and bold text.

"METTL3-mediated m6A modification of MYC mRNA drives temozolomide resistance in glioblastoma." "METTL3 inhibition restores temozolomide sensitivity in xenograft models."

This is the message the scan reader carries. Repeat it in the Q&A preparation.

Decide Panel

This panel carries the deciding experiment — the one figure that, if it failed, the paper fails. Design it to carry the result without the caption.

• One figure, not a composite. A multi-panel figure is acceptable if all subpanels relate to the same result. Composite figures showing unrelated results belong in supporting panels.
• Caption is a sentence, not a paragraph. 20-30 words maximum. The caption should repeat the take-home message in specific terms.
• Scale for legibility. All text in the figure (axis labels, legend) at 14 pt minimum. If a reader cannot read the axis labels from 1 meter, the figure is for the supplement.

Supporting Panels (2-3)

Each supporting panel has one function:

• Panel 2: A validation or control experiment. "We verified this with an orthogonal method."
• Panel 3: A mechanism or extension experiment. "We showed the effect generalizes to patient samples."

The rule: if the supporting panel were removed, would the claim survive? If no, the panel is part of the decide panel. If yes, the panel is optional and should be marked accordingly.

Methods Panel

Methods on a poster are not the full methods — they are a summary that enables the reader to evaluate the deciding experiment.

• One paragraph, 100 words maximum.
• Include: n (sample size), test (statistical test), p-value (or confidence interval), model (biological system).
• Do not include: detailed protocols, reagent tables, the full reagent list. These go to the supplement.

Imply Bar

One sentence on what's new, one sentence on what's next. The new sentence should echo the take-home message. The next sentence should point to a specific future direction.

"This identifies METTL3 as a therapeutic target in temozolomide-resistant glioblastoma." "A METTL3 inhibitor is being evaluated in a phase 0 trial for biomarker enrichment."

Font Sizing

The inverse rule applies: if you can read from a poster standing at the back of the room (5 meters), the poster is legible. If you need to step within arm's reach, the poster has failed.

The Single-Take-Home-Message

The take-home message is the same message as the conference talk: one sentence, one claim, one verb that asserts. If the paper has a take-home message, the poster has one take-home message. If the paper has 12 take-home messages, the poster has one take-home message.

Bad: "We characterized mRNA methylation in glioblastoma." Good: "METTL3-mediated m6A modification of MYC mRNA drives temozolomide resistance." Better: "METTL3 drives temozolomide resistance in glioblastoma."

The take-home message is not the topic; it is the edge of the contribution. Everything else on the poster is support.

Deep-Read Navigation

The 10-minute reader navigates the poster through visual guides. Apply five navigation rules.

1. Numbered panels in reading order. Panel 1, Panel 2, Panel 3. When the reader finishes Panel 1, the number "2" signals the next step.
2. Arrows between panels. Not if the design is already a linear flow (left-to-right in each row, row to row). If the layout is non-linear (e.g., the decide panel is center), use arrows.
3. A flowchart in the methods panel that shows the experimental design in one row. The 10-minute reader checks the flowchart before reading statistics.
4. Color-coded for function: decide panel uses one accent color, supporting panels use a second, methods is grayscale. This helps the scan reader find the deciding experiment.
5. QR code in the corner linking to the PDF. The deep reader can take the paper home.

Q&A Preparation

The poster session Q&A is a 3-minute conversation with standing-room traffic of 10-20 people per session per author. Two preparation templates:

The One-Sentence Introduction

"We found that METTL3-mediated m6A modification of MYC mRNA drives temozolomide resistance in glioblastoma, and METTL3 inhibition restores sensitivity in xenografts."

This is the same one-sentence message as the take-home bar. It can be repeated to every passerby. No modification, no variation — the message should be identical every time.

The Three-Answer Defense

1. "The deciding experiment is the METTL3 knockdown / overexpression experiment in Panel 1."
2. "Controls included the vehicle knockdown, the scramble siRNA, and the in vivo rescue."
3. "The limitations are that this is a patient-derived xenograft model and the inhibitor is a preclinical compound."

If a question cannot be answered in one paragraph, offer to email. Every answer should fit in three sentences or fewer.

Visual Design Principles

1. Three-column or four-column grid. Three or four columns at landscape A0; two columns at portrait or 36x48. Never a single column that forces the reader to scroll up and down the full length.
2. White or light background. Dark backgrounds look striking in photographs but are illegible in fluorescent-lit poster halls. Use white or light gray for the printed poster; use dark mode only for digital displays.
3. Left-aligned text, ragged right. Justified text introduces uneven spacing that strains reading.
4. One accent color, maximum. Use one color (not the full palette) for the deciding experiment. Everything else uses grayscale + that one accent.
5. Consistent margins. 2 cm (0.75 inch) minimum margins throughout. Text that bleeds to the edge is unreadable and unmountable.
6. No gradients, shadows, or texture effects. These print poorly and distract from the content.

Print and Distribution Checklist

Common Failure Modes

A Worked Example: 48x36 Landscape

For a poster on "METTL3-mediated m6A modification drives temozolomide resistance in glioblastoma," layout at 48x36:

+------------------------------------------------------------+
|  TITLE BAR: METTL3 m6A drives temozolomide resistance in   |
|  glioblastoma                                            |
|  J. Author1,2, Author2,1; 1-Univ, 2-Inst                     |
+------------------------------------------------------------+
|  TAKE-HOME: METTL3 knockdown reverses temozolomide       |
|  resistance in patient-derived xenograft models.        |
+---------------------------+-----------------------------+
|  DECIDE PANEL (Panel 1)   |  PANEL 2 (Control)         |
|  [Knockdown rescue]      |  [Western: METTL3 protein] |
|  Caption: METTL3 KD restores| Caption: METTL3 level  |
|  sensitivity in PDX      |  correlates with resistance |
+---------------------------+-----------------------------+
|  PANEL 3 (Mechanism)      |  PANEL 4 (Methods)         |
|  [m6A-seq on MYC mRNA]   |  n=12 PDX, two-tailed    |
|  Caption: m6A peak on MYC |  t-test, p<0.01         |
|  codon 235              |  Student's t-test         |
+---------------------------+-----------------------------+
|  IMPLY: METTL3 inhibition is a candidate combination      |
|  strategy for temozolomide-resistant glioblastoma.    |
|  Acknowledgments: funding, collaborator | QR code      |
+------------------------------------------------------------+

This 8-panel grid (title + take-home) serves the scan at 60 seconds, the deep read at 10 minutes, and Q&A at 3 minutes. The layout scales to 4 columns at A0, or 2 columns at 36x48 portrait.

References

• Loui & Smith, "Talk to the audience" — signposting and engagement rubric, adapted here to poster.
• Nature Masterclasses: Poster Presentation — visual design, scanning vs. deep reading.
• AAAS Communication Toolkit — poster template, 3-minute Q&A structure.
• Anderson, The Way We're Working Isn't Working — TED principles applied to poster: the single idea.
• Alda Center for Communicating Science — improvisation, clarity in the Q&A.

Related Skills

• scientific-communication/conference-talk — companion format for the live talk version of the same content.
• scientific-communication/elevator-pitch — distilled version for the hallway conversation.
• scientific-communication/press-release — the take-home message recast for a general audience.
• scientific-communication/podcast — long-form spoken version for the same research story.