People are lazy. Make it as easy as possible for them to see what you are trying to show. “Mental noise” is my name for ‘work’ your audience has to do before they can actually see what they want to see. In a graph, it can be taking out a ruler to measure a distance, but also tilting the page to read non-horizontal text, or having to look up jargon.

There are tons of ways to reduce mental noise in your charts, but one of the easiest is avoiding abbreviations. Of course, space constraints, general usage, and the specific nature of the abbreviation don’t always make this possible (I wouldn’t replace DNA by deoxyribonucleic acid everywhere), but a specific case where I tend to avoid abbrevations at all costs are countries.

Yes, countries do have official abbreviations — standardized by the ISO 3166 guidelines. There are even 2-letter ánd 3-letter abbrevations for each country. For example, Belgium’s alpha-2 code is BE, and its alpha-3 code is BEL. In most official documents, such as reports by the European Union, the alpha-2 codes are most commonly used. Lack of space is often the main reason, in charts where 27 countries need to be shown.

But these alpha-2 abbrevations are not always as straightforward as in the case of Belgium. Most people will recognize FR, NL or IT. If you know your languages, DE or ES will also make sense. And with some effort, we can decypher SK or AT, even though they don’t always make sense. Some of the lesser known ones are probably EE (Estonia – local name Eesti), HR (Croatia – local name Hrvatska) or EL (Greece – local name Elláda or Ελλάδα).

If you’re looking at 27 EU countries in a list of graph, that’s quite a lot of cognitive load to untangle and recall all of them. Your audience needs quite some background knowledge in order to fully understand your chart. Some people might not even bother looking up country codes they don’t recognize, which of course lowers the relevance of our entire graph. For this reason, I will almost always avoid country abbreviations, and write their names in full. That doesn’t even have to take up much more space:

Source: I encountered a version of this visual on VRT NWS, in an article on the number of political mandates per country (in Dutch). Most of their data was collected from a flash Eurobarometer report “Local politicians of the EU and the future of Europe” by the European Committee of the Regions.

Almost always, gridlines are better than axes. At least for vertical axes.

I admit, vertical axes are the default option, and they have been around for centuries, so they are very well known. But your typical vertical axis also has some downsides. My biggest problem with vertical axes is that they’re often so far away from where the action is really happening. Take a typical chart like this, were nothing is really happening on the left (all the values are zero), but the growth is really big on the right hand side of the visual:

If you want to know the data values near the end of the chart, in 2036 in this case, we almost have to take out a ruler to measure, but the lack of axis ticks (the little horizontal lines next to the numbers) and the distance make that hard to do:

A simple compromise is to move the axis to the right hand side of the visual, where it’s much closer to the ‘action’ — the values we’re actually most interested in:

We had to move the legend to the left in order to free up some space for the axis, but it actually worked out really well. Notice also how we’ve added some explicit tick lines to increase the precision of the visual.

However, moving the vertical axis to the right hand side is not always an option. Often, we’ll have some direct labels or annotations on that side that make it harder to fit in the axis. It would create too much of a barrier between the data and the text. Take this visual for example:

This is a really clean, strong visual thanks to the use of direct labels and some helpful annotations to the right. The only thing I don’t really like is that lonely vertical axis sticking out like a sore thumb at the left side of the visual. However, these labels and annotations are in the way when we want to move the axis to the right:

I’m probably just nitpicking, but that doesn’t look so great to me! In these situations, I will always prefer to switch to gridlines. Yes, they take up more space and create more ‘stuff’ in the visual, but they have two major benefits:

✅ more precision if you’re trying to estimate data values

✅ this precision boost impacts all parts of the visual: left, middle, and right

Here’s how that looks like for the visual above:

I’ve made the colored areas a little bit transparent, so you can still see the gridlines clearly enough. Notice how you can quite easily see that the total value is growing to 200 GW by 2025, and reaching 300 GW by 2030. These intermediate values were quite hard to read in the original visual!

Some final cleanup things we can do:

• nicely align the subtitle and the note with the rest of the visual
• add ticks to the horizontal axis as well
• optimize the annotation to the right, brackets would make more sense here than arrows I think
• add explicit data values for the different categories in 2035 to further increase precision
• move the ‘GW’ label to the tick label

This is how the end result looks like:

Finally, a small bonus tip. If for some reason you’re tight on space, and you have to squeeze your chart a bit to make everything fit, you don’t have to make your gridlines go all the way from left to right. You could consider only having them show up when they’re needed. That would give you some extra whitespace to fit, for example, your title and subtitle:

Of course, that’s something not every #dataviz tool will allow, so that’s only for when you’re willing to make some final custom modifications for your report.

Here’s the full comparison between our original visual, and the reworked chart:

Lots of reports have boring, ugly visuals. Or, even worse, visuals which are really, really hard to understand. The reports by CREG, the Belgian Federal Commission for Electricity and Gas Regulation, are completely different. For example, just browsing through the Monitoring Report — their yearly study on the functioning and price evolution of the Belgian wholesale electricity market — you can immediately see that it’s full of really thought-through visuals and graphs. There are three key principles they apply to bring their visuals to the next level.

Principle 1: add helpful annotations

Sometimes it’s in the small things, like adding a simple arrow to show that the yellow area is equal to the gray area between the grid load and the total load:

The total load and grid load in these curves have very strong seasonal patterns, making it hard to spot whether the delta between them also has seasonal patterns or not. Duplicating that delta at the bottom of the graph — nicely aligned with the horizontal axis — is the best way of making such patterns visual. In this case, there is some seasonality (the delta is slightly lower in winter), but much less pronounced than the seasonal patterns of the loads.

Principle 2: use small multiples to untangle complicated stories

When charts feel complicated, it’s often because they try to explain to many things at once. Different key stories in the data are fighting for our attention. Rather than just eliminating information to make a chart simpler, a helpful technique is to break it up into multiple smaller copies, each of which tells a single part of the story: a ‘small multiples’ approach.

The visual above shows the evolutions for four different countries over a 10-year time period in a small multiples arrangement. Comparing the exact values for the different countries side-by-side is a bit harder, because the bars are a bit further apart from each other than in a traditional clustered bar chart. But comparing the patterns between countries is now easy to do, and that’s the main objective of this chart — to show how different the electricity flow is among Belgium’s different borders.

Principle 3: don’t be afraid to use less common chart types

Not everything has to be a bar, line or pie chart. There are 100+ chart types available to us (if you want an extensive overview, you could check out the Data Viz Project or Data Visualisation Catalogue). Different chart types of course have similar things they can do, but each chart type does have its own strengths and weaknesses when it comes to highlighting certain aspects of your story. If you have an important key message to share, it’s worth considering a few different chart types and choosing the one that shows your message the most clearly.

In the CREG report, you will find bump charts, variable width bar charts (or Marimekko charts if you want to sound fancy), heatmaps, slopegraphs, waterfall charts, and ridgeline plots sprinkled in between the more traditional line and scatter plots. These more exotic charts are added on purpose, with a clear goal in mind, not just to make the report a little bit more fancy (although that is also an effect of chart variety: less boring reports).

Benefits of using better charts

The result of all of this? A 150-page report that doesn’t feel like a chore to read. There is variety, and everything is well explained. Thanks to the clear titles, subtitles and annotations every visual is its own self-contained mini-story — it’s not always necessary to read all the text before and after the figure to understand what’s going on. And most important of all: the graphs are clear and transparent. CREG gets its message across flawlessly, without being hampered by chart clutter, noise, or unnecessary complications. A clear, correct ánd beautiful presentation of the information — that’s what we should all strive for!

Disclaimer

I was (unfortunately!) not involved in the creation of these beautiful graphs. All visuals were created by Senior CREG Advisor Nico Schoutteet. You can read the report on the CREG website.

At Baryon, we love a good podcast—especially when it’s about data visualization. They’re a great way to pick up new ideas, hear how others approach design and storytelling, and stay in touch with what’s happening in the field. Whether it’s a deep dive into visual communication, a fresh take on data exploration, or a candid chat about the challenges of bringing complex information to life, there’s always something to learn (and plenty to get inspired by).

Many shows feature conversations with people doing remarkable work—sometimes the big names you already know, sometimes voices you haven’t heard yet but will be glad you did.

Some of our go-tos are Data Stories, Storytelling with Data, and Data Viz Today, but there’s a whole world of great listening out there.

We’ve pulled together a regularly updated list of our favorite data visualization (and data science) podcasts on Notion. You’ll find them all in one place, with links to Spotify, Apple Podcasts, and Google Podcasts—ready for your next walk, commute, or coffee break.

😶 I don’t like dashboards. There, I said it.

Ok, some nuance: I don’t like _most_ dashboards. The main reason: they’re trying to do everything, everywhere, all at once.

On the spectrum of data visualization, two main clusters of powerful visuals exist:

1️⃣ Data visuals for analysis: useful for data analysts, who have time to explore the data in full detail, with lots of filters, offering many different perspectives on the data. Their goal: extracting the insights from the data.

2️⃣ Data visuals for communication: useful for managers, or a more general audience. They don’t have a lot of time and want to know the major insights, fast, loud and clear. For more complicated stuff, we can craft a strong narrative to guide them through the major insights.

What most dashboards are trying to do, is both of these things simultaneously: raw data goes in, crystal-clear insights come out – or so people expect.

The solution? We make a full-fledged dashboard for the analysts, and a dedicated light-weight version for the management, showing only what they need to know for their decision-making. Or we do our analysis first, and translate those insights into an engaging visual storytelling piece, or an attractive visual report.

As always, we have to think about the audience and their goals. Not just dump the data on top of them, and hope they will figure it out!

🥰 This one is definitely in my top 5 of favourite chart types: a matrix chart!

🤔 What is it? A matrix chart is similar to a heatmap, and it has the same compact design. But rather than relying on color differences, we use bubbles with different sizes to show the data. It is cleaner and less visually cluttered than a heatmap, making it easier to fit in a stylish report design.

⚠️ Potential downside: this chart type works well only when there is sufficient variation between the data points. Otherwise it will be hard to see small differences between the bubble sizes.

💡 Worth noting: not everyone uses the term ‘matrix chart’. Some people prefer ‘proportional area chart’, or in this specific example – because the horizontal axis represents time – a ‘bubble timeline’. I like the term ‘matrix chart’ because it is a visual matrix of data, and it is also how RAWGraphs – my favourite tool to create these charts – calls it.

🐟 Example from the report ‘Toekomstvisie voor de kustvisserij 2024’ (Vision for the future of inshore fishing 2024) that we made for ILVO earlier this year.

Iedereen kent wellicht ‘Kontich’ en ‘Reet’, maar in Vlaanderen hebben we nog veel meer merkwaardige, onverwachte, en vaak grappige plaatsnamen. Heb je bijvoorbeeld ooit al gehoord van Buitenland, Dikkebus, of Grote Homo?

In dit kaartje zetten we de vreemdste namen van Vlaamse gemeenten, dorpen of gehuchten voor jou op een rijtje. Veel zoekplezier!

Ontbreekt er nog een merkwaardige plaatsnaam? Laat het me dan zeker weten op koen@baryon.be.