How to Be Convincing: Three Rules for Creating Persuasive Charts

Creating Persuasive ChartsWe IRI/Nielsen analysts love nothing better than digging through piles of data. But when it comes to communicating our insights through PowerPoint or similar programs, less is more.

To increase the odds of management and customers remembering and following your recommendations, you need to keep charts simple and clear. I’ve developed  three rules for creating persuasive charts:

  1. Limit each chart to one key point. Remove any data that doesn’t relate to that one point.
  2. Use high contrast colors and large font sizes for key pieces of information.
  3. Remove or de-emphasize all other graphic elements. That means no fancy effects, no background colors, no animations and no 3D!

To illustrate, I’ll make over this chart I’ve pulled from my files:

Nielsen data chart before makeover

This chart has great information.  But it’s packed with visual distractions.  That makes it hard to take in its message at a glance.   And one glance might be all you’ll get!

Below, I’ve made over the same chart using my three rules for creating persuasive charts:

Nielsen data chart after makeover

And here’s how I did it, organized by rule:

1.      Limit Each Chart to One Key Point

I started by including my key point in the title of the chart. Rather than “Organic Penetration by Category,” I changed it to “Organic products have captured significant dollar share in fruits and vegetables, with fresh cut salads category particularly well developed.” This title makes the conclusion clear.

I removed categories with repetitive data points. The original chart had nine categories; I’ve reduced it to seven. My chart still illustrates the range of results but requires less reading and has fewer distractions. If it’s important to include data points for all categories, put them in another chart or an appendix.

I also changed the order of categories. The order seemed entirely random before. Now categories are arranged by data value, from highest to lowest.

One thing I didn’t remove was documentation of the data, such as source, time period and geography/market. All charts should include this information. It’s not enough to provide it only at the beginning of a presentation. It may seem repetitive to put it on every chart, but you never know a single chart will be pulled out of the presentation and become separated from this critical information.

2.      Use High Contrast Colors and Large Font Size

I picked a higher contrast color for the dashed line showing the average results. This line conveys valuable information and deserves more emphasis than in the original chart.

The font sizes were good in the original chart. However, I did opt to reduce the font size on the bar labels so I could change them from slanted to horizontal alignment. Horizontally aligned text is easier to read than slanted and worth a slightly reduction in font size.

3.      Remove or Deemphasize All Other Graphical Elements

I changed all bar colors to blue. This is less visually overwhelming, and no information is lost.

I removed the gridlines. Gridlines tend to be overused. They’re often visually distracting and serve no real purpose. When I do use them, I make them a very light gray.

I removed the y-axis labels. Data values are displayed above each bar, so the y-axis labels add nothing except, again, potential visual distraction.

Another change I made (unrelated to my three rules) was to remove the word “penetration” from the chart. “Penetration” has a specific meaning in the context of syndicated data: % of Households Buying. That’s not what’s meant here.  How do I know?  If this was % of Household Buying, the overall food value would not be lower than some of the categories. Instead, I think this chart is showing what % of each category is contributed by organic items.   And it’s probably based in dollars since that’s the typical cross-category measurement. So in my y-axis title, “Organics Share (%) of Category Dollars,” I make clear the data measure used. And instead of “penetration”, I use the word “share” in my title.

My three rules for convincing charts are simple but not always easy to follow. You need to consciously step back from your chart and ask, “What data or design elements can I take away while still making my point?” The more you strip out, the clearer your point.

If you’re intrigued by this topic, dashboard design and visualization expert Stephen Few’s website ( is a great resource. He also provides examples of chart makeovers. His book, Show Me the Numbers: Designing Tables and Graphs to Enlighten, is a classic.

Want to see more chart makeovers? Send me one of your charts and I’ll consider using it in a future post.

Did you like this article? Subscribe to CPG Data Tip Sheet to get future posts delivered to your email in-box. We publish articles twice a month. We won’t share your email address with anyone.

Print Friendly, PDF & Email


  1. Love this, it’s a lost art! My favorite chart in CPG? The classic “bump chart” of weekly volume with price overlay. So many $ driving applications.

    • Sally Martin says:

      Mark, You are so right about the bump chart! There is nothing better for helping you figure out what the heck is going on or at least developing some good hypotheses.

Have a comment or question?