Discussion: Tableau &?Alteryx

I have been quizzed over Data Visualization and Analytical tools a couple of times. Usually, it boils down to choosing between Tableau and Alteryx in one or another form. However, thanks to my work experience that I manage to deploy all Power BI, Tableau and Alteryx in one or another project. While I exploited Power BI and Tableau in two of my former positions, I managed to use Alteryx quite extensively during my tenure of assistantship in the school. 

I planned to write this article after my discussion with a friend, on Tableau & Alteryx, in between the workout sets. We were talking about our experiences with Tableau and Alteryx and the article majorly talks about the pros and cons of Alteryx and Tableau. I am not delving deep into either of the tools. Both tools bring something on the table that others fail to. Moreover, I believe that people should have a good understanding of both the tools so that they can deploy either of the tools depending upon the project requirements. 

After extensively using Alteryx, tonight, I can write that Tableau was the only software in my company that makes the development process a little more difficult than it needs to be. Alteryx is such a baby tool, in comparison to Tableau, that you just have to drag and drop to let the software achieve what you want it to achieve. And, Tableau is more of a kid whom you have to trick to get your work done. No doubt, Tableau is great for 80% of the work that Data Analysts perform on it in their day to day routine but the remaining 20% is something that will slow you down. For that 20%, you have to find a way to fix it. Like a hack to carry out a particular task and unfortunately, that hack will only be able to fix that task. For instance, Fit a moving average on top of the original trend. Trust me, you will be having a hard time getting around this task with common functionalities of Tableau.

“ I hate tableau. It occupies a dark place in my heart reserved for the likes of SAS, SPSS”, wrote a Data Scientist on social media. 

I know so many people that chose ggplot2 over Tableau for building dashboards. And the best part of ggplot2 is feeding it the result of dplyr aggregation operations, so you get everything in one place. 

Anyway, I will not shy away to acknowledge that long before Alteryx was introduced to me, I looked upon Tableau as a fantastic tool. Even now, Tableau is a great solution to be packed with more traditional upstream software and processes that create more of a product for business users that are easily understood. Tableau also helps in generating alerts for Supply Chain related Data Science and Analytics.

Alteryx is a great tool for managing data workflows. A well prepared Alteryx Workbook can easily help you relay the info to your IT team to kick in some ETL tasks. Clubbing with R and Python makes Alteryx the most easy-tool to deploy in its stream. There is a jupyter notebook build in with Pandas, NumPy, Scikit-Learn, etc. 

“ There is an easy( probably not the best) way to integrate python with Alteryx is just to save the code separately and call it using the run command tool. This will let you use any library you want and lets you pull in the existing code without modifying it.”

Alteryx is not just limited to exploiting python but you can use all kinds of Reporting, Parsing, Transforming and Joining tools within Alteryx. You can even use Macros and build the Time Series Forecasting models with few clicks.

Fun Fact: You can use the R module on Alteryx and implement Dijkstra's algorithm, for the slowest path to find along with the network, in merely 7 lines. Ask an IT guy and he will tell you that it will take him a month to do it in R due to spatial preprocessing of data to implement the algorithm. Moreover, you do not have to be an R pro to pull out the jobs using the R module in Alteryx.

And as I mentioned in the beginning that every tool brings something special and lacks something that others bring. Alteryx is versatile but the Alteryx server raises security flags. Alteryx Server is difficult to manage Login services and credentials. The server users under Alteryx are eventually provided access to any data of the workbook and considering my affair with AWS, anything but granular access to data is a horror to me.

Anyway, let us not succumb to confirmation bias and looks at both the tools with a bigger picture in mind. Alteryx is a great tool to be used by Data Analysts but Data Scientists require to build high-performance models which means the usage of Scala or Python on Big Data and Tableau handles big data extremely well in comparison to Alteryx. So, I will not choose one out of the motley of tools until and unless it is necessary because every project exclusively requires to use a specific tool to ease the job. And it is upon you to determine which one you shall go ahead with and hence, it is better to keep your options dynamic to choose the tool that fixes your problem in the easiest possible way without compromising the quality of reports and visualizations.

And go through this article that talks about the best practices of building optimal Alteryx Workflow. 

---------

Rahul Rai Khatri