I've sat through decks where every slide was technically correct and the whole thing still felt like wading through wet sand. Nothing was wrong. Nothing was necessary either. That's what "fluff" usually is — not bad information, just information that didn't earn its slide.

Most advice on this topic tells you to cut words per slide or follow a 6x6 rule. That's a symptom fix. The real problem sits one level up: whole slides, sections, and phrases that exist because someone was afraid to leave something out, not because the audience needed them. Cutting adjectives won't fix that. Cutting the wrong slide will.

Fluff isn't a word-count problem

A ten-word bullet can be filler. A ninety-word paragraph can be essential. The test I use isn't length — it's whether removing the line changes what the audience understands or decides. If it doesn't, it's decoration, no matter how tight the phrasing is.

This matters because people who "trim for conciseness" often just shorten the filler instead of removing it. You end up with a leaner deck that still says nothing. I'd rather see a slightly wordier slide that carries real information than a beautifully clipped one that carries none.

Where filler actually hides

In my experience it clusters in five predictable places, and they're rarely where people think to look:

  • The context slide before the context slide. Teams often build up to their point with two or three slides of throat-clearing — market background, company history, "why this matters" — before saying anything specific. One of those slides usually does the job of all three.
  • Recap bullets that restate the previous slide's title. "As discussed, our Q3 performance showed growth" adds nothing the slide title didn't already say. It's there to sound thorough, not to inform.
  • Qualifiers stacked on qualifiers. "We believe this could potentially represent a meaningful opportunity" is five hedges wrapped around zero commitments. Say what you think happened, or say you don't know yet — both are more useful than the hedge.
  • The agenda slide, when the deck is under ten slides. For a short deck, an agenda slide is a table of contents for a pamphlet. It earns its place past roughly fifteen slides or in a multi-speaker session — below that, it's overhead.
  • Data slides with no stated implication. A chart followed by silence, or by "as you can see, the numbers speak for themselves" — they don't. If the chart doesn't come with a sentence saying what to do about it, it's filler with a legend.

None of this is about tone or polish. It's about whether each element is pulling weight for a specific person in the room who needs it to decide something.

A faster test than "is this necessary"

"Is this necessary" is too easy to answer yes to — almost everything feels necessary to the person who researched it. A sharper question: who in this room needs this specific fact to make their decision, and what happens if they don't have it?

If you can't name the person or the decision, the slide is there for completeness, not communication. Completeness is a writer's goal, not a reader's — nobody in the audience is grading you on coverage. One thing people overlook here: this test also catches slides that exist purely because a stakeholder asked for them in an earlier draft and nobody removed them once the ask changed.

Hypothetical example, not a real client: imagine a ten-slide budget review where slide 4 breaks down a vendor cost by quarter, but the only decision on the table is whether to renew the contract at all. The quarterly breakdown is accurate and irrelevant — the renewal decision only needs the annual total and the alternative cost. Cutting slide 4 to one line frees up room to actually argue for or against renewal, which is the slide nobody built.

Where trimming backfires

I wouldn't recommend cutting everything that repeats. Some redundancy is doing real work:

  • A short recap slide before a break in a long session helps people who stepped out or lost the thread — that's not filler, that's accessibility.
  • For a mixed-language or non-native-English audience, restating a key number in both the chart and the narration reduces misreadings that a native-speaker deck wouldn't need.
  • Executive decks sent without a live presenter often need one extra sentence per slide that a spoken version wouldn't, because there's no speaker to fill the gap. Cutting that sentence to match "in-person" density is a mistake specific to leave-behind decks.

The line isn't "never repeat anything." It's "repeat only what serves comprehension for this specific audience and delivery format" — a recap slide for a live all-hands and a recap paragraph in an emailed PDF are solving different problems, even if they look similar.

A cutting pass that actually works

Rather than editing slide by slide, I go through the deck once asking only one question per pass — mixing passes is how people end up polishing sentences on slides that should've been deleted entirely.

First pass: which slides could be deleted with no one in the room noticing a gap in the argument? Second pass, on what's left: which bullets restate the slide title or the previous slide? Third pass: which sentences hedge instead of commit? Only after those three does a language-level edit for wordiness make sense — and by then there's usually a lot less to trim than you'd expect, because the bulk of the fluff left with the deleted slides, not the tightened sentences.

This also tends to shorten build time on future decks, since a lot of these habits repeat across projects once you know what to watch for.

Structure carries more weight than wording

McKinsey's slide practice is often cited as a design lesson, but it's really an argument-structure lesson — each slide states a single conclusion, and everything under it exists to support that one line. A deck built that way has much less room for filler by construction, because a slide without a clear conclusion doesn't have anywhere to sit in the sequence. If you're rebuilding a deck from scratch, starting with one-sentence conclusions per slide before you build the visuals catches a lot of the padding before it ever gets written.

Visual clutter and content clutter often travel together, too — a slide crowded with decorative elements is frequently hiding a content problem underneath, not just a layout one. If a slide feels busy no matter how you rearrange it, check whether it's actually trying to make two points at once.

Next time you're staring at a deck that feels heavier than it should, don't start by shortening sentences. Start by asking which slides nobody in the room actually needs — that's usually where the real weight is.