One challenge that we are faced with in the Data School (which is extremely good practice for data consultancy ) is to complete a MakeoverMonday in 60 minutes.
For everyone not knowing what MakeoverMonday is: MakeoverMonday (or short MoM) is a weekly Tableau community challenge organized by Tableau Hall of Fame Visionary Andy Kriebel. Each Monday a new dataset is posted, most often alongside a visualization that uses that particular dataset. Your task is to build a new visualization or dashboard, or try to re-build and improve the existing one by the end of the week. There is no right or wrong solution. Instead, it gives you the opportunity to work with new data every single week and practice your Tableau skills.
I HIGHLY recommend it when you are starting out with Tableau as it gives you the practice, practice, practice that you will need. If you are a beginner, you may not want to limit yourself to 60 minutes, but if you are already comfortable with Tableau, feel free to challenge yourself! It is also a fantastic way to try different charts and interactivity features that you have previously learned and build up your Tableau portfolio. The visualizations don’t have to be out of this world, but if you post them regularly, people will see your progress!
As part of our Data School Training, we are tasked with completing such a MakeOverMonday in 60 minutes. This has multiple reasons that carry over to your job as a data consultant down the road:
1.) Learn to scope: You won’t be able to do it all and you have to learn to estimate how much you can get done in a certain amount of time. The only way to do that is to limit yourself to a time and learn over time to estimate how fast you work and how much work you can get done. This will help you tremendously in client projects, placements, and when working with companies. In consultancy, you usually agree on a scope, deliverables, and a timeline before starting a project. So it is important to be able to estimate how much you can realistically do (while accounting for unforeseeable circumstances).
2.) Work under pressure: As a data consultant you will be faced with deadlines. A Friday presentation in front of the CEO, a financial report for a board meeting, a short-notice teaching session, or a short-term consultancy project that needs to be completed by the end of the week. Working with deadlines and under pressure is a skill that can be acquired, but only if you put yourself in challenging and time-restricted situations. MakeoverMondays are a great way to practice those skills in a low-stakes environment.
3.) Practice! No data consultant started out being an expert. No amount of theoretical training will make you an expert. But what will make you amazing at your job is practice (little side note: Contrary to common belief, practice will NOT make you perfect, but it WILL make you better!). The more you practice, the more you feel comfortable with the software, with building certain charts, implementing interactivity features, and formatting. You will get faster at finding the menu settings that you will need. You will apply knowledge that you have previously learned which consolidates that information and you will store it in your medium- or long-term memory.
But how to do it? How to get started?
Admittedly, my first MakeOverMonday in 60 minutes was tough! I had never worked under such a short deadline before, neither in academia, data, or in Tableau. I got very quickly frustrated when something wasn’t working the way I had imagined it, because I felt that I had to make every minute count and every wasted minute would make my end result worse. But, by the second and third and fourth time, I learned to stay calm. It is not a waste of time to take time to understand the data or even to scratch a graph. It is all part of the process to create a visualization that represents the data in the best way possible within the time constraint that you work in.
Here are some lessons I learned and strategies I applied:
1. Create timed sections
I am a planner by heart and I knew immediately that I had to give myself small "waypoints" along the 60 minutes to be able to submit and publish my visualization before the time was over. Therefore, I split up my 60 minutes into smaller sections, each one dedicated to a different task. I will list what worked (and still works) for me below. Note that this is HIGHLY individual and depends on what you are good/fast at. You may be faster at formatting than me. Or you may need more time building your dashboard. I wouldn't recommend taking my times and transferring it to you. Instead, take this as a template and adjust the times to you. Also, it may not work for you at all. You may prefer a different approach to ensuring that you stay within the allotted time.
- 10 minutes: Reading background on data, downloading the data, loading it into Tableau
- 20 minutes: Building charts (including trying out different charts, settling on 1-2 charts and building those out)
- 15 minutes: Building a dashboard (including parameters/filters, title, legend, data source, colors etc.)
- 10 minutes: Final formatting (don't forget to adjust your tooltips!)
- 5 minutes: Upload/publish your work.
2. Choose a subset of the data
When I get a new dataset and I have to build a dashboard in 1h, I immediately look for subsections of the data that I may want to focus on. For example, are there multiple years in a dataset and you only want to focus on one or two years? Are there multiple categories that you could choose from? In a previous MakeoverMonday we had data on paid and unpaid parental leave. Immediately after looking at the data, I saw that this is a potential where I could focus on a subset of the data: I decided to only focus on paid parental leave.
Choosing a subset can also include hiding or ignoring certain columns or variables that you won't be focusing on.
3. Look at data and settle on a story
In the process of choosing a subset of your data and building initial exploratory graphs, you also want to be thinking about a story. What is one insight that you found unexpected? What stands out to you? Which subset of the data shows a big outlier? What results were surprising to you?
If you can answer any of these questions, you already have a story! Then think about what chart fits your story the best and how you want to convey it to the audience.
4. Reduce expectations
This goes hand in hand with being realistic and scoping down. What can YOU realistically get done in one hour? Don't compare yourself to others and think what your coach could get done in one hour or what Lisa could get done in one hour. What can YOU get done?
What is one chart that conveys your story the best? Start with that one chart and if you still have time, you can always add on and make another one. But don't start out thinking that you want to build 4 graphs plus interactivity features and so on. Reducing your expectations will also keep you calmer, which in turn will make you more productive.
5. Be intentional about chart choice and parameters/filters.
You don’t need it all! Often, we aim to implement all possible filters and features that seem useful or we want to show elaborate and complex graphs. But think about what chart conveys your story and the information in your story the best. A Sankey chart may not be necessary if a bar chart conveys your information better.
Also, think about the user of your story. Which filter would the user really need? Could you save time by implementing a filter rather than a parameter?
6. Extra tips:
- Look at previously published dashboards on that same dataset or the topic. Don’t try to re-build but think about the key takeaways of those visualizations and whether anything sparks your creativity.
- Ask for help if you are stuck! You CAN collaborate with others. It is not "cheating". Quite the opposite! You will learn more from asking for help rather than trudging along yourself and then not being able produce something.
- Scratch it before you lose too much time! Know when you have to scratch something. If you tried to build a certain chart for 10 minutes and it is not working, don't get stuck on it. Let it go and try something new!
Let me know if these tips were helpful by reaching out to me on LinkedIn. I love connecting with like-minded data enthusiasts!
https://www.linkedin.com/in/lisa-hitch/
Feature picture by Lisa Hitch