Half-way there!

Wow we made it to the half-way mark for 2017! 26 makeovers done, 26 to go.

We have really enjoyed the past 6 months and it’s been a very successful time for the project and the dataviz community. We’ve welcomed many new participants and have watched with pride as people participated, learned, experimented, grew their skills and their networks within the community.

It’s been fun to see so many different and creatives vizzes come through every week and many of them have inspired others to try different styles and techniques.

I have personally learned a huge amount since joining Andy in running this project and I’m excited about the second half of the year…

This week – German car production and exports

For week 26 we analysed production and export statistics for German cars and judging by the number of submissions the topic must have been interesting or at least easy to work with for a lot of participants.

A number of things stood out for me this week:

Trying something new

You challenged yourself to try something new, be it a new chart type, a different way of telling a story through making specific layout choices, or simply by participating in Makeover Monday for the first time. Here are some examples from the community:

And Thomas had a day of firsts: First Makeover Monday participation, first Makeover Monday live (in Frankfurt), where he also presented his viz as the first person to do so, first tweet and first viz published to his Tableau Public account.

Iteration after iteration

What I enjoy every week, and this week was no exception, is to see people take feedback on board and iterate on their vizzes to make them better.

At the beginning of the year there definitely seemed to be a more competitive spirit of people submitting as quickly as possible and not making changes once their viz was done. Now we see many of you post your viz, open it up to discussion, take on feedback and give each other feedback before refining your data stories and enhancing the visualizations to make sure they communicate the information and insights most effectively.

Not only is it great to see the ongoing conversations and support, but it’s also encouraging for those giving feedback to see recommendations being considered and implemented.

On a side note: we aim to give feedback whenever possible and especially when we notice those receiving it learn from it and improve. What about those who ignore our comments? Doesn’t hurt us, really, but as it becomes time consuming to provide feedback to many people every week, we do aim to pick our battles. If someone doesn’t acknowledge our comments in some way (even if they don’t want to incorporate the feedback, they could still indicate that they’re seen/read it), we will not provide further comments unless specifically asked to do so. There are plenty of people out there who appreciate help and we’d rather work with them :-). So if you feel like you’ve been ignored, my advice is to just let us know specifically if you want feedback and we will happily respond.

This week I noticed Rodrigo going back through his submission to improve it and strengthen his message:

Keeping it simple

Whenever I see a simple, clean and minimalist viz, I am delighted. Day in day out we are bombarded with information and messages through online advertisement as well as on objects and surfaces all around us. We can’t escape it and it’s tiring. So I find it refreshing and soothing to find a simple viz in my Twitter feed. I don’t care if the background is white or black, but when the message is clear and the page is uncluttered, it makes me very happy.

Interestingly, more and more people state the obvious and call out the simplicity of their viz in their tweet. Well, if you are telling me that your viz is simple, then is it really? Because if it is so simple, then surely I will notice that as soon as I click on the image? Andy actually challenged this trend with the following Tweet:

What he referred to is that it’s okay to find datasets challenging and if you want to do ‘just a bar chart’ because you’re not sure how else to approach the data, then that is absolutely fine. There is no reason to put down your own work. We all have times when we’re uninspired, don’t have time or simply don’t find a story. That’s okay.

Keeping it simple can help you because focusing on simplicity is actually quite difficult to achieve. We look at hundreds of submissions every month, a few thousand over the last 26 weeks and we participate every week as well. Of course we sometimes run out of ideas and I had a few weeks lately where I didn’t really know where to take my analysis or my viz. Picking a single story and sticking to a minimalist design helped me create a viz at times when I didn’t feel like it.

But aside from using simplicity as a way to still put something on paper when you lack motivation or ideas, it’s also good to practice having a strong story and impactful visualisations without the need for complex design to carry your message.

As with every other week, there were some lessons learned during ‘Germany week’ (part 1), as we noticed a couple of challenges people faced or mistakes being made, so let’s get into those:

LESSON 1: LACK OF CONTEXT

It is always helpful for your audience to give them some context. Keep in mind that while the #MakeoverMonday community has collectively spent hours looking at the data and are used to working with new topics week in week out, someone browsing through your Tableau Public profile or finding your viz on your blog, LinkedIn, Pinterest or wherever else you share it doesn’t have any idea what the data is about.

Essentially what you’re trying to do is communicate the whole story in a single picture. Tell your audience what they’re looking at and provide a meaningful title as the bare minimum. Often your viz will benefit from a subtitle that provides more detail and maybe a number of annotations relating to specific data points or areas in your charts to call our insights or trends. Providing a link or mention of the data source should be standard and is just the right thing to do. Add information on how to find you (e.g. Twitter handle, blog address, etc.) and your viz is much more complete.

A great example

Michael Sanville‘s viz shows how text, annotations and shading can be used effectively to connect your data with your story to give your audience all the information it needs.

LESSON 2: SHOW ME WHAT YOU PROMISED ME

This week there were a number of vizzes with titles along the lines of ‘the impact of the financial crisis on the German car industry’. But then, that impact was never shown.

Just because your line just plunges in 2008, doesn’t mean your audience will understand that this was due to the financial crisis or that the impact you’re referring to is negative and comes in the form of a drop in production numbers.

The word impact as such is neutral. Could be positive, could be negative. Just tell me what it is. So how about “The financial crisis in 2008 led to a significant drop in car production output in Germany”. This tells me what (the financial crisis) happened, when (2008) it happened, and what the result (drop in production output) was. As the reader of your viz, I know exactly what I’m looking for.

If you promise something in your title, make sure your viz actually shows it clearly. And in reverse, make sure your title accurately describes what your viz shows.

A great example

Alex Jones created a viz titled ‘The Recession caused a slump in German vehicle export levels’. And guess what he’s giving me? A line chart that shows a slump very clearly. He further supports this with a couple of sentences above the line chart, delivering on the headline and therefore on my expectations as a reader.

LESSON 3: TOO MUCH OF A GOOD THING

Colours, icons and images. Many of us like them and like using them. But too much of them does more harm than good.

I noticed a few vizzes this week which had a different colour for each year, resulting in a colourful bar or line chart that is hard to read and make sense of.

Or icons. Do we really need them? I would argue that most people understand the concept of passenger car vs truck perfectly well, so does an icon enhance the visualization?

What about images? Does a picture of a German car really help the reader get to the key insights of your story?

Less is more. As we’ve said many times before, Makeover Monday can be whatever you want it to be and it’s a great playground to experiment and try new things. If you want to go crazy with pictures, all power to you. But there is still a strong case for simplicity when it comes to effectively communicating with data.

If you want me to simply understand that exports as a proportion of total production have developed in a certain way over the years, then focus on that message. Don’t distract me with icons or pictures. And a simple colour palette with one or two highlight colours (okay, make that black, red and golden if you wish), is perfectly sufficient as far as colours go.

A great example

Charlie Hutcheson created a very simple viz that has minimal colour (mostly black, white and grey) and uses a single highlight colour to support his key finding. No icons, no images, just data and clarity. This shows clearly that you don’t need to create complex dashboards with multiple charts and design elements to have a strong story.

LESSON 4: ATTENTION TO DETAIL

Here I will group a few points together.

Typos: Spell-check your vizzes for a more polished look. Whether you’re visualizing simple car export statistics or topics like youth employment in Latin America, please make sure your story comes across clearly and your audience isn’t discouraged by spelling mistakes and other seemingly minor details. They can easily weaken your message and make you look less credible.

If English isn’t your first language, don’t despair. Find a friendly face in the community and send them your draft. Another trick is to read your text from back to front, one word at a time. Or hey, why not just do your viz in your own language? We don’t mind. Just use aliases for your dimension members and you’re good to go!

Watch your aggregations: In this week’s dataset we had production numbers and exports. Exports are a subset of production. That means out of all the cars produced, a certain amount is exported. You cannot add them together as production already includes exports.

If in doubt, simply ask us to clarify questions around the data.

Assumptions can be dangerous – or simply wrong: Be careful when making assumptions based on simple datasets such as this one. What we gave you were production and export numbers per month for passenger cars and trucks. Does that data tell you anything about sales? Not really. It doesn’t tell you how many cars were sold and neither does it tell you when people buy cars in Germany. Yes, it tells you that not a lot of cars are produced during August and December compared to the other month. So what can you conclude from that? Well, Germans probably take holidays during August, but maybe it’s also a conscious choice by the manufacturer to reduce production output during that month. Those are possible explanations, but not definite conclusions.

So be careful with the statements you add into your viz. You could be right, but you could also be way off.

That wraps up the highlights and lessons for this week and now it’s time to look at my favorites…

WEEK 26 FAVORITES

Author: Chris Love
Link: Tableau Public

What I like about it:

• A simple line chart used effectively to tell a story
• A strong title that doesn’t just call out the key insight and provide clues on how to interpret the colours, but also plays with the word ‘stutter’ to reflect the many spikes along the line over time. Subtle and clever!
• Adding filters to the viz proves to me that Chris did a thorough analysis of the numbers to ensure that no matter what dimension members the user chooses, the story still stands. And yes, I did check 🙂
• Using January 2008 as the baseline to compare against provides an interesting perspective
• The colours are well chosen and simple.

Author: Michael Mixon
Link: Tableau Public

What I like about it:

• This is my favourite use of the German colours in any of the vizzes this week. The colours stand out but aren’t overwhelming
• Great idea to use the red headline for the part of the story that really focuses on the challenges the industry faced due to the recession. The other two charts with black and yellow titles have a more neutral message, which aligns with those colours
• Neat alignment of charts and shaded areas for the recession. This helps me to not just focus on the key insight, but with every charge the repeated design further strengthens the message and my understanding of it
• Nice annotations and tooltips, very clean overall design

Author: Adam Crahen
Link: Tableau Public

What I like about it:

• A very clearly focused viz that looks at the cars that were not exported and therefore narrows down the story
• Clear title that tells me exactly what the charts below are showing
• Using a German number plate as a design element brings together the topic and the data but it’s subtle enough to still let the data stand out on its own

Author: Haiping Kuang
Link: Tableau Public

What I like about it:

• The layout works really well with the big numbers, the charts and the alignment of title and text
• Great colour scheme, using shades of red to highlight the impact of the recession
• The highlight table is effective for showing the changes over time in a simple, visual way
• Nice alignment of the charts which means that the years along the highlight table and line charts are all aligned

Author: Robert Crocker
Link: Tableau Public

What I like about it:

• Nice mobile friendly layout
• Great tooltips
• BANs (big ass numbers) work really well for mobile and support the comparison approach of Robert’s viz
• Simple colour scheme and clean charts. I like the addition of the trendline for both, exports and production
• Interactivity to pick the comparison year is engaging and means I will spend more time looking at this viz on my phone, exploring the data and learning more about it