Wow! After an intense first week in the Data School I feel even thirstier for knowledge than when I landed London last Saturday! My curiosity received an unbelievable amount of inputs, and I now have loads of things I want to explore and deepen.
But let’s take it step by step!
The week before this all started was a bit overwhelming for me, as I submitted my Master’s degree dissertation and said farewell to my previous job.
I spent the past five years in a fast growing company that produces great art exhibitions worldwide. As a Business Developer I grew with the company, and I made my mission to let its strategies being always Data-Driven. It was a real challenge: the cultural industry is not the most “Data savvy” (I am being euphemistic…), and my Business Intelligence work was really handicraft!
However, despite lacking of professional BI skills, I developed a specific mindset to let numbers answer questions, and I quickly realized how passionate I felt about this approach!
In the Data School I saw the opportunity to match my passion with professional skills and tools, being trained by some of the best Data Gurus in this world… how exciting! I made my luggage without any doubt, and started this journey!
First, I am grateful to have found a super fun and passionate team to share this adventure with. I strongly believe the diversity of our professional and human backgrounds being really valuable.
Second, I am impressed by the Vision and charisma of our coaches. Their friendly approach is always oriented to maximize the steepness of our learning curve, and this is just great!
Finally, I could not anticipate a more perfect combination of an enjoyable environment and an overwhelming pace!
The very first day we learnt some of the best practices in DataViz, and the second day we had a ‘Tableau Zen Training’ that definitely pushed our limits (and made me really curious!).
Afterwards, we had our first exposure to Alteryx from our favorite Alteryx Ace Chris Love, and started preparing for the Tableau Qualification Exam we will be taking during the Tableau Conference (Argh!).
But all that (which abundantly met my most optimistic expectations) was apparently not enough for our head coach Andy!
On Friday, he managed to introduce us to two of the most visionary and inspiring persons I have ever met: Tableau CEO Christian Chabot, and Alteryx EMEA VP Stu Wilson.
I asked Christian Chabot about the power of Business Intelligence for Small and Medium Enterprises, and he explained his mission of making Data Analysis really accessible to every business. That was the moment I actually realized I was sitting in front of a Disruptive Innovator, who is taking the leads of a true paradigm shift.
In the same way, when Stu Wilson sketched Alteryx’s “Value Chain” on our board, I found myself in front of a second paradigm shift. Indeed, Alteryx enables analysts to transform raw and messy numbers into clean and blended Data, in just a few seconds.
Finally, about the presentation..!
Every week we are going to present a project on a designed topic in front of a multifaceted audience, also broadcasting it online.
My first assignment was a makeover of the second Viz Chris Love has ever made on Tableau: Three Dashboards exploring Food Hygiene Ratings in Local Authorities.
Kindly find the dynamic Vizs on my Tableau Public profile!
- First Dashboard: Best and Worst Local Authorities
This first Visualization is a Dashboard designed for comparing Local Authorities with respect to their Food Hygiene performance. The Visualization is really interesting, but I changed some of its features:
- The Dashboard declares in its title to examine Food Hygiene Ratings by Local Authority, but the first chart I got is a breakdown by Business Type.
- I found it hard to compare Local Authorities using the two lower bar charts. The two charts are built up using a “weighting function designed to punish low ratings” as the ranking criteria, the number of records as the length of the bars and different colors for the ratings. The “weighting function” is “window_sum(attr(6-[Rating])*[Percent Total])”, and it is really smart, but I found it hard to recognize immediately its ranking effect: in fact, I found my eye more likely to compare lengths or colors.
- The use of green and red is not friendly to color-blind people.
I stuck to the aim of the Viz and eliminated the Business Type breakdown. I then changed the colors and opted for a simpler indicator for the performance of LAs: The % of Businesses in the LA that scored the best or the worst rate.
The two charts on the top are about the Best 10 and the Worst 10 LAs, while the two charts below compare all the LAs and can be filtered to compare specific LAs.
- Second Dashboard: Geographical distribution of Ratings
This second Visualization localizes all the rated Businesses as dots on a map, and colors them according to their rates.
- I found the map being overwhelming, until the reader filters it by LA.
- I do not think the ‘Business Types’ bar chart adds to the aim of the Viz.
I re-sized the map to make it the only thing in the Dashboard, and then I changed the colors to let them be more direct, and color-blind friendly.
I added a filter by Business Type and I set it to “Schools”, which is the most interesting cluster.
- Third Dashboard: Correlation patterns
This third and last dashboard tries to explore the correlation between Food Hygiene Ratings and Deprivation, using the English Index of Multiple Deprivation.
- The geographic reference changes from Local Authorities to Lower Super Output Areas, making me feel this Viz to be inconsistent with the previous dashboards.
- Spotting correlations is really hard: the reader is likely to spend a lot of time before getting information about the link between Food Hygiene and Deprivation.
I kept things really easy, just focusing the Viz on the question: Is Deprivation linked to Food Hygiene Ratings? Well, yes, it is. I used Local Authorities as the points on the Scatter Plot, to keep consistency with the other Dashboards. Then I added a label to explain what I meant with “linked” (it is not obvious that the reader knows about correlations and P-Value…). I also added the possibility to filter the scatter plot by Business Type, to see how the link changes.
I am absolutely enthusiastic about this Adventure, I am learning new things at an unbelievable pace, and if I have to summarize this first week using just my major feeling, it would be: I bet the best is yet to come!!