Every deck I get handed to "fix" already followed the rules. Six words a line. Brand colors on every slide. One idea per slide, no exceptions. And it's still not working — the client can't say why, they just know something's off. That's usually my first clue that the problem isn't a rule being broken. It's a rule being followed past the point where it stopped applying.

Most presentation advice gets written as if every deck faces the same conditions: a live audience, a projector, a speaker walking through it slide by slide. A lot of decks don't. They get emailed, skimmed on a phone, forwarded three times before anyone in the room actually sees them presented. The rules that hold up in one context quietly fall apart in the other, and almost nobody writes down where the line is.

"Six lines, six words" is a rule for a room you might not be in

The 6x6 rule — six lines of text, six words per line, nothing more — gets repeated so often it's treated as a law of physics rather than a guideline for one specific situation: a live talk, projected, where the speaker is the primary source of information and the slide is a visual anchor, not the message.

Send that same slide as a leave-behind PDF, and the rule works against you. A reader scanning it asynchronously, with no speaker filling in the gaps, needs the slide to carry the argument on its own — which means more words, not fewer, structured so the reader can extract meaning without you in the room. I've watched teams strip a perfectly clear reference slide down to six words per line for a board deck that was never presented live, only circulated — and the board came back with questions that were already answered in the version before the "cleanup." The rule optimizes for a speaking event; a lot of decks aren't one. If you're building something that has to survive being read cold, it's worth checking how far you can actually push line length and density before it breaks — there's a longer breakdown of where typography rules hold up and where they don't that gets into exactly this kind of condition-dependent advice.

Staying "on brand" can quietly sabotage your charts

This one sounds like it shouldn't need debunking. Of course you use the brand palette — that's the whole point of having one. The complication shows up specifically in data visualization, where brand colors were chosen for logos and marketing collateral, not for encoding numbers a reader has to compare accurately.

A lot of corporate palettes contain two blues that read as nearly identical on a projector, or a primary and secondary color with almost no luminance difference between them — fine for a title slide, a real problem the moment those two colors are assigned to different bars in the same chart. I'd rather see a chart break from the brand palette slightly, using a secondary "data" set of two or three colors chosen for contrast rather than corporate identity, than see two data series the audience literally cannot tell apart because the design brief said "stay on brand." Nobody in the room will fault you for the departure; they will fault you for a chart where they can't tell which line is which. This is closely related to a mistake that shows up constantly in gradient-based charts, where color is doing the audience's math for them and quietly getting it wrong — the way color misleads readers even when it looks polished is worth understanding before you assign brand colors to anything meant to be measured, not just looked at.

Icons don't automatically make a slide easier to scan

Replace a bullet list with icons and a slide instantly looks more "designed." Whether it's more *readable* is a separate question the rule quietly skips. An icon only speeds up scanning if the audience recognizes it faster than they'd read the word — a magnifying glass for "research," a handshake for "partnership." Push past that small set of universally legible symbols and you're asking the audience to decode an abstract shape before they even get to the point it's supposed to represent.

I've sat in reviews where a client insisted on icons for five strategic pillars — none of which had an obvious visual form — and watched the audience's eyes bounce between each icon and its label, working out the connection, which is strictly slower than reading five words would have been. The icon didn't remove a step. It added one, dressed up as design. Icons earn their place when the concept already has a strong visual shorthand the audience shares. For anything more abstract than that, a clean label usually outperforms a symbol nobody would draw the same way twice.

"One idea per slide" breaks the moment the idea has to build

This is close to true often enough that it gets stated as an absolute, and for straightforward decks it's solid advice — cramming three unrelated points onto one slide is almost always worse than three slides. Where it stops applying is a slide that's deliberately building toward a single conclusion across several connected beats: a sequence showing how three inputs combine into one output, or a before/after that only makes sense side by side.

Split that kind of slide to obey "one idea per slide," and you often destroy the one idea it was making — the relationship between the parts was the point. The distinction that actually matters isn't "how many elements are on the slide," it's whether those elements are three separate ideas competing for attention, or three parts of one idea that only resolves once you see them together. The rule is really about competing ideas, not element count, and treating it as a hard number leads people to split slides that were never actually cluttered.

Consistency isn't the same thing as sameness

"Keep every slide consistent" is good advice at the level of margins, title placement, and icon style — a deck that shifts its grid every third slide looks unplanned even when the content is strong. It stops being good advice the moment "consistent" gets read as "identical," and every slide in a 40-slide deck uses the same layout regardless of what it's actually showing.

A full-bleed image breaking up a run of text-heavy slides isn't an inconsistency to fix — it's pacing doing its job. The decks that feel monotonous rather than polished are usually the ones where the designer applied the "stay consistent" rule to layout variety instead of to the underlying structural elements — margins, fonts, color logic — that should genuinely never change. Consistency is a discipline for the invisible scaffolding of a deck, not a mandate that every slide look the same as the one before it. A broader pass through the slide design rules that actually hold up under real audience attention covers this same tension from the layout side, if you're rebuilding a deck that's drifted toward either extreme.

Why this keeps happening

None of these five rules are wrong, exactly — that's what makes them stubborn. Each one is genuinely correct for the specific condition it was written for: a live audience, a spoken narrative, a designer filling in context the slide alone can't carry. The formatting itself isn't even where most of the time goes wrong, either — a global PowerPoint study by the software firm empower, run with GfK, found that formatting alone eats up roughly 37% of the total time professionals spend working in PowerPoint, and a meaningful share of that is spent enforcing rules that don't fit the deck in front of them.

The fix isn't memorizing five exceptions to bolt onto the five rules. It's asking, before you open the software, what kind of object this deck actually is — presented live or read cold, one idea per slide or a sequence that builds, brand-consistent or measurement-accurate — and letting that answer decide which "rule" applies this time. The rule was never the problem. Applying it without checking first was.