6. Receive Feedback and Iterate
After you’ve built an initial dashboard, it’s time to get feedback from users and iterate. If your dashboards are going to be used by a wide audience, pick a few of the users you trust to give you feedback. Pick the people who will be frank and honest with you. This isn’t the time to get pats on the back. This is a chance to get constructive criticism so your product is as polished as it can possibly be.
Have you ever read a story about someone finding a meteorite? It probably contained a fantastic story about a smoking space rock careening across the sky, striking the Earth, creating a big, smoking crater and the finder being there to miraculously witness it all and dig it out. It then gets valued at hundreds of thousands of dollars and the finder walks away wealthier.
Imagine you are working on the following dashboard:
You’re pleased with where it’s at overall, but you’ve run into a dilemma. The dashboard helps answer the overall question “Where should we invest our marketing dollars?”, but it’s very hard to compare individual states. Maps are great for high-level geographic overviews but poor for comparing individual values.
Thank goodness! In Tableau Desktop 2018.3 Tableau officially released the “button object” for use in dashboards. Historically, you had to create a separate worksheet and go through a complicated process to connect multiple dashboard via navigation buttons. Not anymore!
Traditionally the set feature in Tableau is used to create a subset of predefined values. Those may be handpicked (i.e. How do these 5 products perform regionally?) or they may be chosen conditionally (i.e. Where are our top 100 customers by revenue located?). They are generally used to create predefined values you can filter on.
When you embark on a Tableau dashboarding project, you are creating a new product. When Apple releases the newest iPhone, they aren’t putting out a rough draft. They’ve done extensive user and product testing to make sure its the best product possible. You can (and should) use the same design sprint methodology on your own projects to ensure success.
When you embark on a Tableau dashboarding project, you are creating a new product. When Apple releases the newest iPhone, they aren’t putting out a rough draft. They’ve done extensive user and product testing to make sure its the best product possible. You can (and should) use the same design sprint methodology on your own projects to ensure success.
I was working with a client recently and needed to do find and replace in a Tableau calculation. While Tableau does have a native FIND() function, it just finds if a character string exists and tells you what position that string starts at. Not particularly helpful when it comes to replacing.
Preparing survey data for analysis in a data visualization tool is notoriously difficult. It's not as simple as being able to perform a single pivot or filter.
My grandparents have a beach house in Island County and every 4th of July there is a big parade and community get together. One of the events is the “Penny Hunt”. The adults scatter a bunch of coins (of varying denominations) in the sand for the kids to search for. As kids, my brother and I got fed up with blindly digging in the sand so we convinced our dad to get us a cheap metal detector. I remember pulling in $40 the first summer we put it to use. Not bad for a couple of kids.
A client recently challenged me with the question, “How could I show a bar chart in Tableau that shows values for the top 10 displayed as 10 individual bars but group all the rest of the values into a single bar called “other”…
Let me be completely honest with you. I am writing this post in the aftermath of the 2019 Rose Bowl where the University of Washington lost to Ohio State so my views and opinions are most certainly biased. It’s not a surprise to me that the Huskies lost. I think Ohio State was the better team. Rather, it was a decision in the 4th quarter by the Huskies that surprised me…
Nope.
If you need to pivot data from a SQL table for use in Tableau, Tableau Desktop’s default pivot feature can’t help you.
You might never have heard of a self join but you might need one. Occasionally, tables in a database are structured so that it makes sense to join a copy of a table to itself…
Sometimes you need to use data in Tableau that isn’t in a clean, denormalized format. It might have been exported from an application or prepared by a coworker in a way that Tableau doesn’t like.
Totals in Tableau are notoriously rigid. You can’t add two totals lines; one for summarized values and one for averaged values, in the same worksheet. You can have one, the other, or allow Tableau to use a field’s default aggregation for totals.
