2026-06-28
A pre-flight checklist for decks that face partners, boards, or clients
Fifteen checks that catch the mistakes senior reviewers spot in the first thirty seconds — sources, fit, parallelism, and the so-what.
Senior reviewers don't read your deck the way you wrote it. They scan for credibility signals in the first thirty seconds — and a handful of mechanical mistakes will cost you the room before your argument gets a hearing. This is the checklist we use.
The story
- Titles alone tell the whole argument (the horizontal-logic test).
- One idea per slide. If a slide makes two points, it's two slides.
- Every content slide answers "so what?" — in the title or a takeaway.
- The recommendation appears early, not as a slide-30 reveal.
The numbers
- Every figure has a source on the slide, not just in your head.
- Estimates are labeled as estimates, with the basis stated.
- Numbers that should reconcile do reconcile (totals, bridges, CAGRs).
- Chart types match the message: composition → stacked bars or pies,
bridges → waterfalls, trends → lines, exact values → tables.
The mechanics
- Nothing is clipped, nothing overflows, nothing almost-aligns.
- Bullet lists are parallel: same grammatical form, comparable length.
- Fonts, colors, and spacing are consistent across every slide.
- The deck survives PowerPoint: send it to a colleague and have them edit
a chart. If they can't, fix your export path.
The room
- The deck works without you narrating it — assume it gets forwarded.
- The appendix carries the depth; the main line carries the argument.
- Someone hostile has read it before someone friendly presents it.
Most of the mechanical items on this list shouldn't require human vigilance at all — layout fit, consistency, sourcing, and editable exports are properties a well-built tool can guarantee. That conviction is why we're building Slidechat the way we are. The story checks, though, remain your job — and they're the fun part.