A client once handed me a spreadsheet with fourteen columns and asked for "one clean slide." Not a summary slide, not a section — one slide. My first instinct was to start deleting columns. That instinct was wrong, and figuring out why changed how I approach every dense dataset since.

The problem was never that the data was complicated. The problem was that nobody had decided yet which question the slide was supposed to answer. Cut columns before that decision is made and you're just making an arbitrary version of the same overwhelming table — smaller, but no clearer.

Simplifying isn't the same as shrinking

Most people treat "simplify the data" as a subtraction problem: remove rows, drop decimals, shrink the font until it fits. That works right up until the moment someone in the room asks "wait, where did the Q3 number go" — and you realize you deleted the one data point that mattered to them specifically, because you didn't know it mattered to them specifically.

What actually works is deciding the single claim the slide needs to support, then keeping only what defends that claim. Everything else doesn't get deleted — it gets moved. Into an appendix slide, a speaker note, a backup chart you pull up only if someone pushes back. Nothing that could matter to someone in the room should disappear entirely; it should just stop competing for the front-row seat.

The chart type is a decision, not a default

I wouldn't recommend opening PowerPoint or Keynote and reaching for whatever chart type you used last time. A stacked bar chart and a line chart can display the exact same numbers and lead an audience to two different conclusions — one emphasizes composition, the other emphasizes trend — and if your underlying point is about trend, a stacked bar will quietly work against you even though the data is technically correct.

Here's where it gets less tidy than most advice admits: sometimes the "correct" chart type for the data is the wrong choice for the room. A scatter plot might be the honest representation of a noisy dataset, but if the audience has thirty seconds and no statistics background, a scatter plot reads as noise, not insight. In that case a simplified trend line — clearly labeled as smoothed, not raw — often communicates the real finding better than technical accuracy does. That's a trade-off, not a shortcut, and it's worth saying out loud in the room rather than hiding it.

Layering beats flattening

One thing people overlook: audiences don't actually want less data, they want sequenced data. A dense table flattened into a single view forces everyone to build their own mental hierarchy in real time, which is exhausting. The same information staged across two or three reveals — headline number first, breakdown second, outlier explanation third — lets people absorb it in the order you intended instead of the order their eyes happened to land.

This is where three-dimensional or layered chart formats can genuinely earn their place instead of just looking impressive. In my experience, cube-style or stacked visual formats work best specifically for hierarchical or staged data — cube-based chart templates are built around exactly this kind of layered reveal, where each face or level carries one part of the story instead of all of it competing for attention at once. Used on a single flat comparison, the same 3D structure adds confusion instead of removing it. The format has to match the shape of the decision the data is making, not just the shape of the data itself.

Tables aren't the enemy — unstructured tables are

There's a reflex in presentation design to treat tables as the thing you simplify away from, as if any grid of rows and columns is inherently a failure of design. That's not been my experience. A well-structured table — clear grouping, one visual anchor per row, consistent alignment — can communicate a comparison faster than a chart, especially when the audience needs exact figures rather than a general shape. The failure mode isn't the table format itself; it's a table that was built by dumping a spreadsheet range onto a slide without deciding which rows deserve emphasis.

When I do use tables for dense comparisons, I lean on pre-structured table layouts rather than building the grid from scratch, mainly because the grouping and alignment decisions are already made — which frees up the actual working time for deciding what belongs in the table, not how to draw its borders.

The one-number rule, and its exception

If a slide is meant to persuade rather than just inform, it should generally carry one number the audience can repeat back five minutes later. Not zero numbers — vague reassurance without data reads as evasive — and not twelve, which reads as either padding or distrust of your own argument.

The exception is dashboards and status reports, where the entire point is that different people in the room need different numbers, and picking one to headline would misrepresent the report as an argument when it's actually a reference document. In that specific case, a well-organized multi-metric layout does the honest job that a single highlighted number can't — the goal shifts from persuasion to fast lookup, and the layout should shift with it.

Where geographic data breaks the usual rules

Regional or geographic datasets deserve a separate mention because the "simplify by cutting" instinct actively backfires here. Cut a region out of a map to reduce visual noise, and you haven't simplified the data — you've silently changed the scope of the claim, which is a different kind of problem than clutter. If the slide is about performance across a specific set of markets, the map needs to show exactly that set, no more, no less, with color coding doing the simplification work instead of omission. Editable regional map templates — focused Central America map layouts, for instance, rather than a full-world map with everything outside the relevant region grayed out as an afterthought — keep the scope honest while still letting color and overlay carry the comparison.

What I'd tell someone building this slide tonight

Before touching fonts or colors, write down the one sentence the slide has to prove. If you can't write that sentence, no amount of chart-type selection or white space will fix the slide, because the actual problem hasn't been solved yet — it's just been dressed up. The data was never the hard part. Deciding what you're willing to leave out, and being honest about why, is.