It's the season of Iron Viz, and this year's theme is "LOVE" and it's anything you're passionate about! It could be a personal health journey, a musical group, movies, games, hobbies or anything that makes your heart full. And like we always say here in #GamesNightViz...
You gotta viz what you love!
If you have already started on your entry but want some help in iterating your story or design, join the community-led feedback sessions organized by Sarah Bartlett. These feedback session will do wonders for getting you back on track, and making sure you are presenting your best work ever! Also, Tableau has released a comprehensive judging criteria this year that is very clear on what is expected from each of the pillar - Analysis, Story and Design; which is great news for both new and returning Iron Vizzers!
But of course, the GNV team's not going to leave you stranding too! Will, Tina and Louis got together to share some of their useful tips on each of the judging criteria so you can use it as a guideline to craft your perfect #IronViz entry!
Before I jump in, there are many supportive materials out there, some of which Will and Louis have also shared below, but I also wanted to highlight Sarah Bartlett’s piece. In this article, Sarah covers details (and resources) regarding the criteria and scoring rules that were released this year which are important to review and understand.
Disclaimer: The points I am highlighting are pitfalls I myself have fallen into. No one is perfect and that is why participating in Iron Viz is about learning and setting different goals for yourself to achieve, regardless of if you place in the competition. I learn something new about myself every year I participate, and at the very least I hope you can take away some notes for yourself too.
Create a *Tableau First* design.
The first piece of advice I can offer is remember that this is a Tableau competition. Meaning, if you design a beautiful dashboard, but used Python for 90% of your data manipulation and Figma for 90% of your layout and copy, and Tableau for 10% in each of those categories, you are unlikely to score very well in Iron Viz criteria. You could certainly submit it for other competitions, but Iron Viz is about Tableau, so you should focus your execution on being predominantly Tableau, and using other tools for 10%, instead of 90%. Save the inverse for a different dashboard;, it will still be awesome!
Set constraints for yourself.
If you find that you’re adding things to your dashboard with the intention of making your audience go “wow!”, chances are you’ve added too much. Stick to the basics, finish designing it, and then tinker with your story, identifying what additive things contribute to your narrative. It isn’t about the “wow”; it's about what contributes to your narrative, further engages your audience, and increases the impact of your visualization. Simplicity can help you with this. Furthermore, creating some boundaries for yourself can be helpful. For example:
- Pick a color-blind palette before you start your design
- Select an accessible font and only use that font throughout the entire dashboard
- Pre-define your narrative arc and number of visualizations you want your dashboard to have
These things will help you with scope creep, as well as maintaining cohesion throughout your visualization.
Stick to a design process for yourself...and don’t stray from it!
I’ve presented about Design Processes in the past, but there is some truth to sticking to your process. For Iron Viz 2023, I completely abandoned my process and it showed in my end product, as well as the amount of chaos I created for myself. This year, I’m forcing myself to stick to my process and not jumping ahead to the visual design element until I finish the data collection and mockup process. It is painful for me to not be able to design right now, but it is also helping me dramatically in solidifying my intentions of what my end result will look like and the themes and narrative I want to pull through. It is also going to be such a relief when I get to the design phase and I’m done with the data, yay!
Designing with accessibility in mind.
This is a big one. I would argue it’s almost as important as the *Tableau First* design, because there are so many ways to do this easily, and so many ways to ignore it unconsciously. At the very least, make sure your colors and font selections are accessible. There are a ton of resources out there that let you view images from different types of color blindness, as well as resources that assist with picking the right font (another topic I’ve covered previously). Accessibility is about cultivating an inclusive environment, and ignoring it is likely going to score you low.
Storytelling is key to successful design;
But I’m not going to cover this because Will/Louis will below!
Lastly, if you’re not having fun anymore, stop.
An old saying (and truth) was when someone would say they could “hear you smiling through the phone”. This illustrates how contagious energy can be. The same goes for your design. If you hate your topic, don’t like your design, and don’t like the constraints, then it is going to show up in your end product, whether you realize it or not. Iron Viz takes time away from your life, and if you are feeling this way about it, then it might not be for you, and that is totally valid! One of my favorite mantras for myself is “go where the energy takes me”. If I’m not feeling it, I’m done.
As always, if anyone has questions, or is looking for feedback, feel free to DM me on Twitter @covellicreative.
Design and storytelling go hand in hand to present your data in a way that resonates with your audience, but in order for you to truly nail that down - you cannot cut corners on your analysis. Like the evergreen phrase “Garbage in, Garbage out”, the accuracy and depth of your analysis builds that “Why should you care?” to your audience.
I love how Tableau expanded on the judging criteria for this year’s Iron Viz, and it is absolutely critical to building that outstanding IronViz submission. Here are my five tips to help your visualization hit all the notes and stand out.
Don’t be afraid to collect your own data
With a theme like “Love”, it can be really daunting to find the right dataset especially if your topic is very niche. But don’t let that hinder your passion, as we live in a world of data. Here are some scenarios where you would find your data.
Games
- Check how many games you own and note down the game details like production date, title, number of players, estimated game time, how many times you’ve played it
- Reviews from other players (Metacritic, Boardgamegeek, IGN, Gamespot) that you can manually scrape data into your own google sheet. If you’re unfamiliar with web scraping, you can definitely check out tools like browse.ai or octoparse
- And of course GNV archives 🎮
Fitness
- Check your Fitbit, iWatch or any smartwatch and there should be an option to export all the data from your phone to an email address of your choice.
Others
This isn't a catch-all, but you can definitely look for more. If you have very specific data needs, you can reach out to us and we can see how we can point you in the right direction, or possibly link you up with someone who can! 👍
Now, from the expanded judging criteria for Analysis this year, there's essentially three key concepts.
Data quality and consistency
You should ensure that your data is error-free, and that should be something that comes as a byproduct of doing data cleaning. Making sure that:
- you truly understand every dimensions and measures in your dataset
- data is of verified sources if you're not using your own personal data
- values are properly formatted and cleaned (i.e. numbers have no text in it, removing trailing or leading spaces from text)
- standardizing the data values (i.e. GamesNightViz, gamesNIGHTviz, GAMESNIGHTVIZ are all the same, pick one and stick to it)
Data accuracy
Above all, the data should represent the data.
- Custom calculated fields should be tested for accuracy, especially when you work with percentages and LOD functions.
- Axis should not be truncated to give a false sense of impact
Analysis Depth
The main role of a data visualization is to be able to communicate deeper insights of the dataset to your uninitiated audience. That said, with a theme like "Love" and if you are taking on a topic that you know a lot about (for example: a specific game title - like what I did for last year's Iron Viz), assume the audience know nothing about it and build context for them.
Throw in your 1-min elevator pitch and use data to convey that point. In the IronViz entry I did last year "It was NEVER about the Dragons", I started out by stating outright what the game is about (i.e. Inspired by the greatest fantasy stories) and adding additional data points to build relevancy. Maybe my audience don't know about Dragon Age, but they may know about Baldur's Gate, or Lord of the Rings, or Game of Thrones.
Then, add more context to your data. Is the metacritic score for Dragon Age considered good or bad? And by adding a simple benchmark reference line, it build that visual analysis that anything above is good. Analysis and design goes hand in hand, because as data viz practitioners, we have to use colors, fonts, shapes and tap into pre-attentive attributes/Gestalt Laws to "force" your users to see the same insights as we do. Don't settle for defaults. Every text, color, image and viz should be designed intentionally. When in doubt, you can definitely use the J.A.R.V.I.S. framework that I playfully coined a while back.
If you need a refresher on cognitive theories, you can check out the virtual meetup I did for the Singapore TUG back in April. Or check out the blog post I did after.
But having said that, you can really enhance your story by building more context for your audience. One rule of thumb I always tell my mentees is, explore your data and find 10-20 interesting pieces of information. Continue to refine that into 3-5 key statements and have supporting data facts to add weight to those insights. That way, you can really sell the story.
If you need a quick run-thru on how to explore your data and craft a meaningful story from there, you can check out the recent Live Builds I did on Esports.
Tableau-first
Lastly, Tina touched on this above, but one of the key criterion for the Iron Viz is making sure your insights and analysis are produced within Tableau. If you'd like, you can definitely do exploration on Excel or Python, etc; but when you're building your visualization, make sure those insights are recreated in Tableau too!
Crafting a compelling narrative is crucial to creating an outstanding IronViz submission. Here are five key points or concepts for storytelling that can help your visualization stand out
1. Clear Objective
What are you trying to tell us with this visualisation? When receiving your work overall your dashboard should have a clear topic or question it's trying to answer, and each piece of your dashboard contributes to the topic/question. Be aware that you may not know what this message is when you start your project and explore the data but it will emerge over time. Steve Rawling, speaks about the "curse of knowledge" where we are too absorbed in our own idea to explain it to someone else. Imagine you’ve found that interesting insight in your analysis, how could you share it without confusing your audience? Steve identifies three key approaches to consider:
Change
- what does your data show about a change/threat,
- or how are you using this data to make your own changes
New Information
- when the data makes you realise something new,
- or it made you spot something that isn’t right
Personal Benefit
- what can this data help people achieve,
- or where does this data help solve a problem
The Science of Winning by Varun Jain gives an analysis of thousands of chess matches but what’s in it for me? Well in the end Varun gives us a winning strategy for how we can improve our chances of winning in chess. Here there’s a clear personal benefit for the audience.
2. Engaging Hook
The hook, it's what gets your reader interested to explore your viz versus all the other distractions after their attention. "10 unexpected habits of cats", "What your accountant doesn't want you to know about...", many examples of these can be found all over news sites, social media, and YouTube. Importantly if you're going to include a hook, you need to deliver on it, otherwise, the user leaves disappointed and unrewarded for their time. In Storyteller Tactics, Steve Rawling studied story hooks from watching around 1,000 TED talks, and found these distinct hooks:
- Questions
- Unexpected
- Ironies
- Relatable
- Knowledge
- Superlatives
Looking over 2023's Iron Viz Gallery I can find a few hooks in dashboard titles.
- Questions: HOW DID DUOLINGO BECOME MY FAVOURITE GAME? by Deborah Simmonds
- Unexpected/Ironies: It was NEVER about the Dragons by Louis Yu
- Relatable: Going For Broke - The Friendship Ruining Power of Monopoly by Ann Pregler
For designing your hook, step back and look at your data story and think:
- What questions are you answering?
- What is something unexpected from my analysis?
- If it's a more niche topic how is this relatable to everyday life?
3. Structured Flow
After you’ve analysed the data, asked many questions, and found different insights it's about bringing this all together in a structure that someone new to the work can easily follow. When telling you’re data story I’d first advise you to work in threes, namely the 3 Act Structure.
Considering this type of structure helps give a good flow to the analysis you’re presenting. It also helps identify areas that have too much, or too little content, as well as areas that are diverting away from the endpoint of the story.
- ACT 1 - What’s it all about? What are our metrics? What’s the problem we’re trying to solve?
- ACT 2 - Why does everything in ACT 1 matter? What happens if the metric changes? Why should your audience care?
- ACT 3 - What should we do? What’s the outcome of all of this? What action should we take?
You can also think of this as: What? So What? Now What?
Organising the beats of a data story was an important consideration for Paul Ross, 2023 Iron Viz champ, when I caught up with Paul he spoke about how he created a structural flow with his Iron Viz feeder entry.
Lastly, one piece of advice I love about arranging your story beats or sections of analysis comes from the creators of South Park. The advice is that between your story points should be a “but” or “therefore”, if it's “and so this happened” your reader can become uninterested and disengaged with your work.
4. Context and Explanation
16,000.
On their own numbers hold no significance to us. Here I’ve put 16,000, currently, it relates to nothing and has no meaning to me. Say if I change this to…
16,000 people fall in love each day.
I can now relate to the number but still, I don’t know whether this is a good or a bad thing, are more people falling in love than before? Adding a historical component, e.g. Up 10% since 2010, brings me some positivity that more can find love in this world. Context is key for making the data accessible to your audience and more impactful. You should be thinking why would someone care about this metric?
When I was exploring the caffeine content of expressos in high street coffee chains I was surprised by how greatly the values differed. Pret had the strong expresso but why should I care about 180mg, well after a quick Google I’m able to say drinking more than 3 of these expressos could cause health risks. My user now knows that more isn’t always better and maybe more careful about which coffee brands they visit and how much they consume.
5. Emotional Connection
We relate to people and personal stories. In this presentation, Hans Rosling is talking about global energy consumption but brings up a personal story of his Grandma and a washing machine. “Throughout her life, she had been heating water with firewood, and she had hand-washed laundry for seven children. And now she was going to watch electricity do that work.” It is one data point as part of the wider story but puts into context what the data means on a personal level, painting a vivid picture in your mind.
In the 2022 Iron Viz final I needed to add that personal story to my work, I had seen how education levels had changed over time but how did that change people’s day-to-day lives? I created three personas about people around the world who would have grown up in these areas and how learning to read and write impacted their lives and the people around them.
If you like to learn more about storytelling with Storyteller Tactics, then check out these articles to turn data into an exciting story:
- https://pipdecks.com/pages/whats-it-about
- https://pipdecks.com/pages/story-hooks
- https://pipdecks.com/blogs/storytelling/data-detectives-card-of-the-month
Lastly, in most cases you have a story within your data somewhere, these tips are here to bring out that story and present it to your audience (and those judges!). For example, you've analysed your data:
- What have you learned from your data that you didn’t know at the start? Build this into a Story Hook
- What’s the data trying to show? What are the good and bad sides of metrics you’re looking out? Why should the audience care about what happens next? Storyboard these answers into a Structural Flow
- How does it impact us the audience? Could you give us a person's situation we could relate to? Give us that Emotional Connection
And that's all we have for you today 😉
Hopefully this is useful to help you pivot, refine and strengthen your IronViz submission! Best of luck, and looking forward to seeing everyone's work! 📊📈
