Voronoi Creator Resources

Data storytelling is a growing field that is constantly evolving to reach new audiences and find novel ways to help people understand the world better. This page is dedicated to sharing knowledge on creating content on Voronoi, and will be a work in progress as we add resources.

Creating Mobile First Content

Traditionally, the bulk of data-driven work has been done on laptops and desktop computers. It’s where people crunch numbers and experiment with data visualization. As a result, most data storytelling is formatted for larger screens. This runs counter to how the majority of internet users now experience content, which is to say, through their mobile devices.

Here are some quick tips on designing data visualizations to be mobile-first:

Prioritize Key Information

  • Simplify your visuals to highlight the most critical information. On smaller screens, less is more—avoid clutter or unnecessary details.
  • Use a hierarchical structure to emphasize the takeaway. For example, bold the key figure or headline, and place supporting details below it.

Optimize Layout

  • While Voronoi supports longer content, we recommend designing within a 1200x1500px canvas for optimal performance.
  • Stick to a vertical layout that’s easy to scroll through on mobile devices, or break horizontal designs into a slideshow.

Minimize Text and Use Mobile-Friendly Font Sizes

  • Reduce reliance on long paragraphs or labels. Keep your copy succinct. 
  • When designing on a computer, keep in mind that text size will appear roughly half as large on mobile devices. For example, 30pt on your computer will display as 15pt on a phone.
  • A good starting point for type sizes are:
    • Minimum body copy and labels: 25pt 
    • Sources and footnotes: 23pt

Test Readability in Real Contexts

  • Preview your designs at actual mobile sizes. The best test is to review it on your phone. 
  • Check colors: Colors can look different across screens. To ensure your visuals are both high-contrast and accessible, test your color choices on multiple devices, especially on your phone.
  • Type sizing: Check that your text hierarchy is adequate for mobile viewing.

Choosing Your Data Sources

Choosing the right sources is an important element of building trust with the Voronoi community. As you’re gathering research, here’s some criteria you can use to choose the most appropriate source.

The Organization

How credible is the organization that has produced the data? This is the most important criteria when choosing a source. Here are some things to look out for:

  • Original source: It’s common to come across an article that is quoting the original source. Secondary sources, such as media articles or blog posts, can be helpful when you want to quote interpretations of data or better understand a topic. However, make sure to drill down to the primary source to do your due diligence. Wherever possible, we suggest using data that is freely available online.
  • Credibility: Is this a trustworthy source? You can help determine this by asking yourself the following questions: Does the source have a track record of producing good quality data? Are they a respected authority in their field? Do respected media organizations such as the New York Times/Wall Street Journal quote them? What does their “About Us” page or Google search results tell you about the organization?
  • Bias: Aligned with credibility, we recommend using non-biased sources wherever possible such as non-partisan groups, think tanks, and government organizations. You can also use tools such as Ground News to see left/centered/right coverage of the same story or Mediapoly to check the bias of articles on Twitter.
    • In some cases it may work to quote a biased source, such as when we are highlighting two sides of an argument. We recommend clearly addressing any potential bias (e.g. “the Republican representative said…” and “the Democrat representative said…”) in these cases. The sources tab is a great place to post this type of contextual information.

One example of a common, respected source is the Federal Reserve (U.S. Central Bank). The Fed is an original source that produces its own data and it is freely available. It is a well-known, trusted source that makes decisions independent of the government.

Timeliness of Data

How old is the data, both in terms of the published date and the “as of” date? If you are addressing a current topic, we suggest using data that was published within the last two years. That said, the topic at hand strongly influences how much recency of data collection matters.

  • When older data is more acceptable: Some sources, typically international or government organizations, produce data on a less regular or lagging basis. For example, the OECD’s data on air pollution only extended up to 2019 as of 2023. As well, some data is only collected in multi-year intervals (e.g. censuses). Some trends, such as demographic shifts, occur more slowly, so projections from previous years can still be useful.
  • When recent data is more appropriate: If market conditions are changing quickly, data that is 1-2 years old may no longer be relevant. For example, many 2019 reports on topics like foreign investment were no longer relevant in 2020 once COVID-19 dramatically impacted markets. Stock market or cryptocurrency data may have a shelf life of only a few days.

The best case scenario is when data is used shortly after it has been published.

Cohesiveness of Data

Whenever possible, use one source rather than putting together a dataset based on multiple separate sources. This is because different sources may use different methodologies or even contradict each other.

Methodology

The criteria a source has used to calculate their numbers is another critical point for judging the reasonableness of a dataset. Do the underlying calculations do a good job of reflecting what is being measured, and do the assumptions seem to reflect reality?

There are a few common methodology mistakes. For example, a source that measured trends during a very small or atypical time period, and then applied those insights to the market at large is sometimes referred to as “cherry picking” data. 

In the case of surveys, paying attention to the sample size is important. As well, look for clarity around the type of people surveyed and the methods used to survey respondents.

Story and Framing

Uncovering the story in a dataset starts with asking: “What is the most interesting, useful, or surprising insight?” Look for patterns, trends, or outliers that stand out or spark curiosity—this becomes your key takeaway. Once identified, craft a narrative around it, ensuring the takeaway remains front and center in the design process. By keeping this focus, you’ll make more informed design and storytelling choices, creating visuals that are both impactful and meaningful.

Data to Viz

Found a great dataset but unsure how to begin? Here’s a list of tools that creators can use to transform raw data into visualizations:

Flourish, Data Illustrator, and RawGraphs allow you to download your visualizations as SVG files, making it easy to import them into design software. A common tool used to refine graphics and to add in additional layers of storytelling is Adobe Illustrator.

Choosing the Right Chart

Not sure which chart type to use? Here are some resources to help you understand different visualization types and how to use them effectively: