Launch, grow, and unlock your career in data

December 22, 2022

How to Calculate Customer Churn Rate in Tableau

No comments

In this tip, I’m show you how to calculate the 3-month and 6-month churn rate in Tableau. To calculate churn rate, all you need to do is:

1. Create 4 calculations

2. Follow the 8 simple steps to build the view

Customer churn rate is the % of customers that you’re losing in a predetermined time period. Churn rate is bad because if you don’t know how many customers are leaving, then you won’t know the impact on your revenue and you would be able to make changes to reduce churn.

December 13, 2022

#MakeoverMonday Week 50 - Which Countries Get the Most Vacation Days?

No comments
This week's data set focused on the vacation time and paid public holidays (or lack thereof) of countries around the World.

I remember when I started working at Coca-Cola. I got 10 days of vacation time for my first 2 years. Then I got 1 additional day for each year thereafter. It was terrible, but unfortunately the norm in the US.

I started by trying to create a tile map of the World, but it was going to take too much manual renaming of the countries than it was worth. So I got the idea that I wanted to show rankings, which led me back to Workout Wednesday Week 31. The view was pretty simple to create; follow along in WatchMeViz below.

My final viz is below the video, or you can view it here.

December 8, 2022

How to Conduct Customer Retention & Cohort Analysis

No comments
In this tip, I’m going to show you how to analyze customer retention and conduct cohort analysis in Tableau.

With cohort analysis you group your users based on a shared characteristic to better track and understand their behaviors. Cohort analysis is typically used to understand customer churn or retention.

Cohort analysis is nearly always done for an app launch.
  • First, down the view, the users are divided into cohorts based on when they first installed the app
  • Then, across the view, the users are tracked for 10 days after the launch to see who continued to use it.

Comparing these two groups allows us to see when users are churning so that we can then investigate why.

December 6, 2022

#MakeoverMonday Week 49 - How good are Americans at perceiving demographics?

No comments
I really liked the original visualization by YouGovAmerica this week, so during Watch Me Viz, I spent the hour working to replicate it. 

Easy to replicate:
  • Barbell chart with dots on the ends and a line connecting them
  • Coloring the lines by the larger value
  • Choosing the colors
  • Formatting the view (except the labels)

Challenges/things that should be easier:
  • Labeling the dots/ends of the lines; this took several iterations on the calculations
  • Creating the color legend as headers; I had to use reference lines, turn on totals, move the totals to the top, then work on the formatting to get them close. They still aren't perfect.
  • Creating the horizontal reference lines; this required me to create a dummy axis

While there are only 3 challenges, I estimate those three challenges took 75-80% of my time. I can only imagine how much more productive I could be if formatting didn't take so long.

In the end, I'm satisfied with how close I got to recreating the original chart. I like the additional context I added by coloring the lines connecting the dots.

If you missed #WatchMeViz, check it out below. Below the video is my final visualization.

November 22, 2022

How to Create a Full Year Calendar Heatmap in Tableau

No comments

Calendar heatmaps are a great way to visualize time-series data over an entire calendar year. Individual days are shaded on a light-to-dark gradient based on a measure that has been added to the visualization.

Creating a heatmap for a single month in Tableau is pretty straightforward. It gets a bit more complicated when you want to view an entire year at once. In this tip, I going to show you, step-by-step, how to create a full year calendar heatmap in Tableau. The workbook is below this video to download.

#MakeoverMonday Week 47 - Worldwide Railroad Infrastructure Quality Rankings

No comments
This week was a fairly simple data set. Rankings by year by country for a single measure. One thing I found tough to get my head around was the 1-7 scale of the quality ratings. I converted the values to a 1-10 scale instead.

I had limited time for #WatchMeViz, so I quickly went through a few techniques you will find interesting:

  1. Recreating the original bar chart
  2. How to create a bump chart
  3. How to use a diverging color palette and alias country names
  4. How to create a ranked dot plot
  5. How to use parameter actions to sort the view
  6. How to use a filter action to remove the highlighting

This definitely ended up being the largest viz I have ever created (1300x3200). You can view the dashboard below the video. Click on the image to view the interactive version (or click here).

I hope you found it useful. If you did, please give the video a like and either share this post or share the video so that you can help others...pass it along. :-)

November 17, 2022

How to Create a 3 Level Scatterplot Drilldown

No comments

If you want to drilldown multiple levels in a scatterplot, this tip is for you. In this tip, I show you have to create a 3 level drilldown with a scatterplot.

Naturally after I showed how to create a 3 level bar chart drilldown, someone immediately asked how to do it with a scatterplot. Unlike the bar chart drilldown, the scatterplot will only show you marks for the current level. 

For example:

  1. When I’m at the highest level, I see only regions.
  2. When I drill down one level, I only see States within the Region selected
  3. When I drill down one more level, I only see postal codes that are within the region and state I selected.

This example shows you how to drill down from Region to State to Postal Code. To use this for yourself:

  1. Download the workbook
  2. Copy all of the calculations into your workbook
  3. Swap out the dimensions

Note that you will need to recreate the sets and set actions. This will work with any combination of 3 dimensions.

November 14, 2022

#MakeoverMonday Week 46 - The Cost of Cocaine & Heroin

No comments

Quite the interesting topic this week. I get an email from The Economist regularly and there was an article about the wholesale price of cocaine. That led me down a rabbit whole and I found the data, but also for heroin and also for retail prices.

On Watch Me Viz, I stuck with simple lines charts that compared the price for a country to the overall price for all of the countries in the data set. I allow the user to choose a drug and a country.

I then wanted to show how to Dynamic Zone Visibility feature works (if you haven't seen it, be sure to watch it back). In the end, a simple dashboard that compares the measures in two ways, allows for some filtering, looks good, and that's it.

Check out the dashboard below the video.

November 10, 2022

How to Add Hundreds of Custom Color Palettes

No comments

Tableau comes with some great color palettes that have been designed to work well together. And you can add your own custom color palettes by making a couple of simple changes to the preferences.tps file that comes with Tableau Desktop. For example, you may have a color palette that matches your company’s brand.

In this video I show you how to add 3 different types of color palettes to Tableau:

  1. Categorical
  2. Sequential
  3. Diverging

You can add as many color palettes as you’d like. While you can have as many colors as you want in each palette, the Color shelf only displays 20. Here's a link to my color palettes.

Once you’ve updated the preferences file, restart Tableau and they’ll all be there for you. Below the video is the code you need to create each of the color palette types. Make sure when you copy in the double quotes that they are straight double quotes and not angled.

Categorical Palette

color-palette name="Palette Name" type="regular">

Sequential Palette

color-palette name="Palette Name" type="ordered-sequential">

Diverging Palette

color-palette name="Palette Name" type="ordered-diverging">

November 8, 2022

#MakeoverMonday Week 45 - Who Americans Spend Their Time With

No comments
I found this week's data set super interesting, but also super sad in some ways, like people are more and more alone as they age. Or people spending less time with their partners when they are much older, clearly showing when people typically pass away.

I found the original visualization quite good, so during #WatchMeViz, I spend the hour replicating it. You can see during the video how much time formatting takes, and how I got annoyed about a new bug with colors. In the end, I go pretty close to the original, but some of the highlighting, and particularly the speed of the interactions couldn't be replicated in Tableau; Tableau simply isn't as responsive.

My viz is below the video or check it out here.

November 4, 2022

Tableau Charts: Comparison to Budget & Prior Year

No comments

Are you looking for ways to improve how you visualize comparisons between 2 to 3 performance measurements in Tableau? 

In many performance monitoring dashboards, you need to make comparison of actual values to budget or target. In this video, I will show you effective ways of making these comparisons. 

I will also show you how to compare your performance to prior year (PY) to see if performance is improving year over year. 

One thing you'll notice in these examples is the consistent use of color. Make sure your metrics use the same colour throughout your dashboards, reports and worksheets. The colors should be easy to differentiate. 

I have 4 designs that you can apply to make any comparison look better and easier to understand. 

Download the workbook for a bonus 5th chart type that compares these three metrics over time.

November 1, 2022

#MakeoverMonday Week 44 - Fundraising vs. Spending by Members of the 117th Congress

No comments

The midterm elections in the US are next week. If you're able to vote, please do. Democracy is at stake. 

This week's data was about fundraising, spending and debt by people currently in Congress. If you missed #WatchMeViz, I showed 16 different ways to visualize this data set. Hopefully they give you a bit of inspiration for creating your own.

Catch up with the show below. In the end, I went with a bar chart that compares funds raised vs. spent by State. I also have a gantt bar to show the difference between the two. I have a tutorial of that chart here.

Below this video is an image of the dashboard I created. Click on it to see the interactive version on Tableau Public.

October 27, 2022

Which chart type should you use?

No comments

I get asked quite often how I know which chart to choose and when. There are lot of times now when I don't really think about it, but that because of my experience. When I started, I had no idea which chart would work best for my data.

To make it easier for you to choose the right chart type, here's a list of resources I use when I need to know which chart to use and when. 

Financial Times Visual Vocabulary

Use this guide to help you decide which data relationship is most important in your story, then look at the different types of chart within the category to form some initial ideas about what might work best.

Chart.Guide Chart Chooser

Link  |  The charts are split into different categories, based on their primary use. You can then pick a chart type within each group. I particularly like the icons that tell the primary use for the chart: analysing, communicating, monitoring, and confusing.

The Data Visualisation Catalogue

Link  |  This is similar to You can see all of the chart types alphabetically or organize them by function or as a list.

The Tableau Chart Catalog

Link  |  Collated by Kevin Flerlage, this workbook provides a list of chart types and links to them. 

The Data Viz Project

Link  |  This is by far the most extensive set of charts. Each chart includes a detailed description of the chart type and its functions. There are also options for viewing the charts by chart family, data input, function and shape.

The Visualization Universe

Link  |  I really should be using this resource more often. It's curated by some of the best data visualization experts in the world. You can sort by most popular, alphabetical, and which chart types are trending.

October 25, 2022

How to Create a 3 Level Drill Down in Tableau

No comments

I've showed you in several videos and WatchMeViz live streams how to create a 2 level drill down. And naturally the follow up question is always, how can I go one level deeper? This tip has exactly what you're looking for. I show you how to create a 3 level drill down in Tableau.

This example shows you how to drill down from Category to Sub-Category to Segment. All you need to do for your situation is replace the dimensions I'm using with yours. This will work with any combination of 3 dimensions.

Check out the interactive version below the video.

This type is inspired by this example from Lindsey Poulter -

October 24, 2022

#MakeoverMonday Week 43 - How have Major League Baseball Pitch Types Changed?

No comments

This week's data set was pretty simple. We had 21 baseball seasons and a column for each pitch type. Pivoting the metrics made it much easier to work with for me as I could then split the view by pitch type.

During #WatchMeViz (below), I create a trellis view, showed how to create groups, sets, set actions, sparklines, LODs, custom number formatting, creating a mobile view, and more.

Thank you for tuning in. Here's the video and below is my visualization.

October 17, 2022

#MakeoverMonday 2022 Week 42 - Food & Drug Administration Inspections in America

No comments

Having looked at the original visualization this week (here), it was pretty clear to me that the creators desired to create a KPI style dashboard. In a way it's good that I picked up on it that way. However, it wasn't designed particularly uclarly well to communicate most efficiently.

I decided during Watch Me Viz that I would stick with their idea of a KPI dashboard, but design it in a way that I thought would be easier to understand.

As per normal, it took way too long to faff about with the formatting. I made a couple of small changes after I finished the live stream. Click on the image below the video to check it out. Leave a comment on the video on YouTube if you have any questions. (Comments aren't appearing on the blog for some reason).

Have a wonderful week!

October 14, 2022

How to Use Show/Hide Buttons in Tableau Dashboards

No comments

Show/Hide buttons in Tableau allow you to do just that, show or hide containers, worksheets, blanks, text, legends, even the title. You can now hide ANY object on a dashboard.

One of my favorite use case is adding a show/hide buttons for a filters container. This helps you hide all of the filters at once and increase the space for your charts.

Keep in mind that it's a good idea to name every object in your dashboards in the Layout section so they are easier to identify when creating the buttons.

October 11, 2022

#MakeoverMonday 2022 Week 41 - UNDP Human Development Index

No comments

From UNDP:

The Human Development Index (HDI) is a summary measure of average achievement in key dimensions of human development: a long and healthy life, being knowledgeable and have a decent standard of living. The HDI is the geometric mean of normalized indices for each of the three dimensions.

This week, I really liked the original visualization, so during #WatchMeViz,  I spend the time recreating the visual. I got most of the way there in an hour and finished it off in the evening. The questions on the live stream are super helpful...thank you!

A couple things I learned:

  1. Hover action interactivity is very, very slow in Tableau. I ended up changing it to a select action.
  2. Labeling is overly complicated.
  3. Tables in a tooltip can look really good. And they are very responsive.

The final visualization is below the video. Thanks for watching!

October 5, 2022

How to Create a Starburst Chart

1 comment

Starburst charts, sometimes called sunburst charts or circular timelines, are one method of displaying data over time. They may look cool, but they aren't exactly easy to understand. If you are going to use them, I'd advise you to check with your audience first to ensure they understand how to read them.

In this video, I take you through how to build a single starburst as well as a series of starburst charts as small multiples or trellis chart.

All of the calculations are in the description for the video here. You can download the workbook below. 

October 3, 2022

#MakeoverMonday 2022 Week 40 - Income Inequality Around the World

No comments
Of course, as soon as I posted the data set this week, the UK PM decided to backtrack on the tax cuts she has promised before. In my opinion, this is a good. But it kinda made that part of the analysis irrelevant.

Still, though, I was curious to see how the UK compared to the rest of the World. During Watch Me Viz (below), I started by rebuilding the original chart, which I quite liked. I then looked at the data over time, but it was quite sparse and difficult to do any meaningful analysis of.

So I decided to stick with a single chart that looked like the original, but it includes all countries and some filtering options.

To learn how I approached the analysis and built the charts, watch the video below. My final dashboard is below the video.


September 27, 2022

Parameter Actions: Geographical Drill Down in Tableau

No comments
Thanks for watching this video about performing a geographic drill down in Tableau!

In this video I will walk you through:
- How to drill down from Region to State
- How to get back to the State level
- How to create the parameter and parameter action

Download the workbook here.

September 26, 2022

#MakeoverMonday Week 39 - Average Internet Speeds Across Europe

No comments

My goal this week was to correct my failures from last week, so I focused almost exclusively on building the starburst/sunburst chart. As promised to the viewers, I will also record a separate tip video for how to build them.

Here's the WatchMeViz and the viz is below that.