It is ten at night and the deck for tomorrow’s review is done: the company template, a stock photo of a handshake on the title slide, and a chart whose colored legend sits in the far corner. Nothing on it is wrong, exactly. But on each slide the audience will spend its first seconds working out where to look, while you are already talking.

Slide design, for a non-designer, is mostly about attention: helping the audience see your point with the least effort while they listen to you. You do not need an eye for layout, only a few principles from research on how people learn from words and pictures together:

1. Cut what does not serve the point (the handshake photo, the logo on every slide)
2. Point the eye to what matters (one highlight color, everything else gray)
3. Put related things together, in space and in time (labels on the chart, bullets revealed as you reach them)
4. Show key words, not your script
5. Make the headline the point (a sentence, not a topic)
6. Keep the look steady and readable from the back

Nerves and rehearsal are a separate problem, handled in the beginner’s guide to public speaking; this article stays with what goes on the screen.

Slide design is about where the audience’s attention goes

Slides work or fail through the audience’s working memory, the small mental space where people hold and combine what they are seeing and hearing. Richard Mayer’s cognitive theory of multimedia learning, reviewed by Mayer himself in 2024, rests on three ideas: people process words and pictures in separate channels, each channel’s capacity is severely limited, and real understanding needs active work to select, organize and connect the material.¹

While you talk, your audience is already using its verbal channel to follow you. A slide full of sentences asks that same channel to read at the same time, and a slide cluttered with logos and a distant legend asks the eyes to search before they can understand. Each extra item takes a share of a small budget, and short-term memory holds only a few items at once.

Take a quarterly results slide. If the audience must match five colored lines to a legend in the corner while you explain which region grew, some will still be matching colors when you move on. Label each line at its end and the chart asks for nothing but understanding.

What this means for you: judge each slide by the effort it asks of a listener who is also following your voice, not by how full or finished it looks.

Most slides break the basics, and taste does not catch it

Common slides break basic principles of perception and memory almost everywhere, according to a 2012 study led by Stephen Kosslyn, then at Stanford University. His team checked 140 PowerPoint slideshows sampled at random from the web, from academic research to business, against rules built on eight psychological principles, and not one slideshow was free of violations.²

The usual faults were overload and poor legibility: whole lists shown at once, long bullets, too many items per list, and text too small or set entirely in capitals, italics or bold. Business decks did no better or worse than academic ones. When the team surveyed 205 people who regularly sat through presentations, the fault they rated both most common and most annoying was a main point buried under irrelevant detail.²

A third study hints at why good intentions are not enough. Shown pairs of slides, one following a rule and one breaking it, 48 volunteers picked the better slide about four times in five, yet for about one in six of those right answers they could not say what made it better.² Recognizing a clear slide is a different skill from building one.

Your own deck is the hardest to judge: you already know what each chart is meant to show, so it reads clearly to you; the audience sees it cold, for only as long as you stay on the slide. The sample dates from around 2008, when the study began, and the rules came from Kosslyn’s own 2007 book on presentation design, but the practical lesson holds: check each slide against explicit rules, like the six below, rather than asking whether it looks fine.

Further reading

• 

  Storytelling with Data: A Data Visualization Guide for Business Professionals

  by Cole Nussbaumer Knaflic

  A practical guide to decluttering charts, labeling lines directly and using one highlight color, the same edits this article recommends.

As an Amazon Associate WiserHours earns from qualifying purchases.

What the research on words and pictures finds

Design choices that reduce needless mental effort improve how much people learn from slides and other multimedia, according to a 2022 overview by Michael Noetel and colleagues at the Australian Catholic University, which pooled the findings of earlier meta-analysesmeta-analysis: A study that combines the results of earlier studies on the same question into one overall estimate. Pooling makes the estimate more precise, but it cannot repair the studies it pools: a meta-analysis of surveys is still survey evidence.Full entry in the glossary. Placing related words and pictures together and cueing the eye to the key point were among the most effective design choices, and design mattered more when a presenter set the pace, as in a lecture, than when learners could go at their own speed.³

The study

Moderate evidence

Words plus pictures, tested in 1,189 studies and summed up in 2022

Across the reviews, 11 design principles improved learning. The largest benefits came from captioning videos for second-language learners, from putting related material together in space or time (for example, labels on a diagram rather than in a separate legend) and from signaling, meaning cues such as arrows, highlights or a pointer that show where to look. Design principles mattered more with complex material than with simple material, and more when the pace was set for learners, as in a lecture, than when they set their own.³

The pacing result matters most for presenters. A colleague reading your deck at a desk can linger on a crowded chart until it makes sense; a room watching you present gets that chart only while it is on screen, and is listening at the same time. So the version you present live deserves a stricter edit than the one you send round afterwards. The main caveat is that the outcome was learning from lessons, by pupils and students; we found no study measuring whether cleaner slides win more budget approvals or sales.³

Applying slide design principles to a real deck

The six edits below apply those findings to an ordinary work deck. They run roughly from the best-supported to the least: the first three are backed by meta-analyses of experiments, and the last leans on standards and rules of thumb.

1. Cut what does not serve the point

Remove anything that does not help the audience understand the point of the slide, however attractive it is. Researchers call these extras seductive details, and in a 2020 meta-analysis summarized in the Noetel overview, leaving them out improved learning, especially when they stayed on screen rather than appearing briefly.³ Eye-tracking studies in the same overview showed why: the extras pulled attention away from the core information.³

Myth

An interesting photo or a funny image makes a slide more engaging, so it can only help.

Fact

Interesting but irrelevant extras draw attention away from the point. In experiments, learning was better without them.

In a work deck, seductive details are the handshake photo, the icon beside every bullet, the logo on every slide and the background texture. Each has to earn its place by helping the point: a photo of the product being discussed earns it; people smiling at a laptop do not.

2. Point the eye to what matters

Tell the audience where to look. In the Noetel overview, signaling, meaning cues such as arrows, highlighting or a pointer, was among the best-supported principles, and eye-tracking studies showed that these cues helped learners focus on the key details.³ The overview’s practical advice is to use color on purpose, for example one contrasting color for the thing that matters.³

In practice, make the whole chart gray and only the bar you are discussing blue, circle the number you want remembered, and dim the parts of a slide you have finished with. Without a cue, the audience spends attention searching; with one, the search takes a moment.

3. Put related things together, in space and in time

Place each label next to the thing it describes, and show each point when you talk about it, not before. The Noetel overview found large benefits for this principle, which researchers call contiguity: labels placed on a diagram instead of in a separate legend, and words shown at the same time as the matching speech.³ Its example of the timing version is familiar to any presenter: revealing one bullet at a time as you speak to each.

1. A headline that states the point: a short sentence instead of a topic label
2. Labels on the lines: no trip to a legend in the corner
3. One color for the point: everything else in gray
4. Decoration removed: the picture did not help the point, so it went

Schematic, not real data: the lines show the layout, not any result.

Kosslyn’s team found the opposite habit almost everywhere. Showing a whole list at once, instead of building it up point by point, was the most common rule broken in their sample.² Position does work of its own: people see items that sit close together as a group, and read a label as belonging to the nearest element, which Kosslyn’s team calls perceptual organization.² Keep each label beside its data and unrelated items apart, and the layout explains the structure before you do.

4. Show key words, not your script

If a slide has no useful picture, put a few key words on screen rather than nothing, and never your full script. A 2012 meta-analysis by Olusola Adesope and John Nesbit, mostly with college students, found that people learned more when spoken words were accompanied by the same words in writing than from speech alone, but the advantage showed up mainly for material without pictures and at a pace the learner did not control. Across studies, showing key terms taken from the narration was linked to better learning than showing the narration word for word.⁴

Here the research is more nuanced than the popular advice to put almost no words on slides. Key words give listeners an anchor when there is nothing to look at; a pasted script makes them read and listen at once. For a budget update with no chart, “Travel down, hiring up, net flat” on screen should work better than both a blank slide and the paragraph you plan to say. Where you have a meaningful chart or diagram, the Noetel overview suggests speaking to it rather than relying on text.³

5. Make the headline the point

Replace topic titles such as “Q3 results” with a short sentence that states the slide’s message, such as “Q3 sales grew in every region except the north.” That is the core of the assertion-evidence approach, which Michael Alley at Penn State has written about since 2005: a sentence headline stating the main message, supported by visual evidence rather than bullet lists.⁵

In a 2013 experiment co-written by Alley, 110 engineering students were randomly assigned to watch a technical presentation with one of two slide sets: assertion-evidence slides, or slides following common PowerPoint defaults. The assertion-evidence group’s essays showed better comprehension and fewer misconceptions, and the group reported lower mental effort, though immediate multiple-choice scores did not clearly differ.⁶ The two sets differed on six design principles at once, not only the headline, and the study was funded by Alley’s own Penn State center, so it supports trying the approach rather than treating it as settled.

6. Keep the look steady and readable from the back

Change the look of a slide only when the meaning changes. Under Kosslyn’s principle of informative change, audiences expect any change in how something looks to carry information, so a color that means “our product” on one slide should not mean “competitor” on the next.² One template, used consistently, does this for you.

Then check that the back row can read it. Kosslyn’s team treated text smaller than 20 point as too small to read easily, a heuristic rather than a tested cut-off.² For contrast, the web accessibility standard WCAG 2.2 asks for a contrast ratio of at least 4.5 to 1 between normal text and its background, and 3 to 1 for large text, noting that people with low vision often struggle to read low-contrast text.⁷ It was written for web pages, not slides, but it is a sensible benchmark. A quick test: shrink your slides on a laptop screen and see what you can still read.

WiserHours

Before you present

• Each slide has one point, and nothing on it is only decoration
• One highlight color marks what matters on each chart
• Labels sit on the chart; list items appear as you reach them
• Key words on screen, not the script you will say
• Headlines state the message as a sentence
• Colors keep one meaning throughout, and text is readable at laptop size

Find this checklist online:
https://wiserhours.com/public-speaking/slide-design/#checklist-slide-design-before-you-present

Where the evidence for each slide rule is firm, and where it is thin

Meta-analyses of experiments back cutting irrelevant extras, cueing attention and placing related material together; the headline advice rests on one small study, and type size and contrast on standards and rules of thumb.

The bottom line

Judge a slide by one test: can the audience see its point without effort while listening to you? Cut what does not serve that point, cue the eye to the one thing that matters, label things where they sit, and put key words rather than your script on screen. The research behind these principles comes mostly from classroom experiments, but the reason they work, a small and shared attention budget, is the same in a meeting room.