My post "Color your way to visual finesse" is selected as the "Best of the Tableau Web in October 2014" by Tableau Software:
http://bit.ly/best-tableau-web-october-2014

This follows an earlier recognition in August 2014:
http://bit.ly/best-of-tableau-web-august-2014

Thanks, Tableau community!

Your death depends on where you live. This is not surprising. We learnt it in things like "Gun, Germs and Steel" the book. Now we have serious data to prove it.

Recently, a viz of the day showed the causes of death in USA. Andy Kriebel published a makeover of the viz. As a follower of Andy's blog, I noticed there is a pattern in the distribution of the death rates from Andy's viz. It inspired me to further look into the data from a geo perspective. From the study, I got a better understanding of the distribution of the major causes of death in the US.

This Tableau story is here to present the analysis process as well as the visual tools for further exploration and analysis. The findings are presented in the summary. It may lead to more questions like why there is huge difference in annual death rate between some states. We allow you to visualize those differences.

The main points are:
- Leading causes distribution on a map
- Leading causes distribution by longitude on scatter plot
- Leading causes distribution by altitude on scatter plot
- State rank by death rate- Cause rank by death rate
- State rank by cause- Cause rank by state
- In-state percentage per cause comparison
- Top 2 causes' contribution 
- Summary

The original data is from CDC 2011. The rate is per 100,000 population.

Following the steps in this video, I can create a funnel chart.
http://www.tableausoftware.com/learn/tutorials/on-demand/funnel-charts

But, there are a few things that the trainer didn't elaborate in order to make a complete funnel chart in practice. So here you go.

1.Get data in order

Before creating a funnel chart, you need a simple table with 2 columns: Phase (dimension) and Value (measure). This step is required especially if you have a dynamic data source for each phase.

In our example, we assume there are 4 phases as follows:

Phase
Visits
Leads
Opportunities
Funded

Just copy the above 5 rows (1st row being the column name) and paste it into a blank worksheet in Tableau. You will have one dimension: Phase.

In practice, the values for each phase may come from the aggregation of different calculations or different data sources. For example, one is from Google Analytics and another is from Salesforce. We need to create a calculated field to calculate the values for each field.

Right click on the dimension "Phase" to create a calculated field "Value". The expression will be something like

if attr(Phase)="Visits" then sum(Google_Analytics.Visits)
elseif attr(Phase)= "Leads" then sum(Salesforce.Leads)
elseif attr(Phase)= "Opportunities" then sum(Salesforce.Opps)
elseif attr(Phase)= "Funded" then sum(Salesforce.Funded)
end

Note that we need to use attr() because otherwise, Tableau will complain that we can't mix non-aggregate and aggregate values in an expression.

So the above is the pre-requisites for a funnel chart with dynamic data.

2.Sync dual axis

At a certain moment in building the funnel chart, we need to use dual axis and synchronize both axis. However the synchronization is not straightforward. Actually we need to convert all values to Float for both axis to sync. This seems not documented anywhere. I found this workaround in this thread:
http://community.tableausoftware.com/thread/109657

So, the previous expression becomes:

float(if attr(Phase)="Visits" then sum(Google_Analytics.Visits)
elseif attr(Phase)= "Leads" then sum(Salesforce.Leads)
elseif attr(Phase)= "Opportunities" then sum(Salesforce.Opps)
elseif attr(Phase)= "Funded" then sum(Salesforce.Funded)
end)

3.Fix a graphical glitch

There could be some annoying glitch showing up in the funnel chart. See the chart below and compare them. You will see the glitch in the chart on the left between blue and green.

This can be remedied by re-ordering measures in "Measure Values".
4.Label properly

The usual labeling approach won't work for the funnel's phases. One can't just drag and drop Phase and Value into the Label shelf, because the funnel is a complex chart that has got multiple components. You will see multiple labels at the same spot if you use the conventional labeling. The trick is to use "Mark in Annotate". Right click anywhere in the area of interest and select Mark in Annotate.
Now we will be able to edit the labeling text and variables in Mark.


By adding up the above steps, you will get a beautiful funnel chart!

*All the data used here are fictional.