


Recently, I was invited to do Alpha and Beta testing of Tableau Prep and I am excited to share with you what Tableau Prep is doing (as of Beta version 1.0) for data management. I’ve been impatiently waiting to get my hands on it ever since. Click & Tweet!Īs a Tableau Consultant, I’m excited to know that the amount of time I spend doing data prep and cleansing is about to end, allowing more time for the good stuff - analysis! Tableau is Changing Face of ETLĪt the 2016 Tableau Conference, Tableau gave a demonstration of an ETL tool called Project Maestro (as of release in April 2018 it is called Tableau Prep), which they were in the process of developing. Some studies show that data analysts spend 80% of their time cleansing and prepping data and just 20% on actual analysis. And once tomorrow’s, next week’s, or next month’s data arrives, you have to do it all over again. Tableau would be the only programm running on that server (the database is remote).Admit it, in order to analyze your data, you’ve spent hours porting data into Excel to fix/modify it, filter out irrelevant, bad or test records, enhanced the data by adding additional fields from other data sources using vlookup, added calculated fields, etc. My situation would be that it would be running on a virtual server with decent but not unlimited memory resources. My only worry is that, when reading a bit about this online, some reviewers have mentioned that Prep can be unusably slow when dealing with anything more than example data sets. I like that I can just send this straight into Tableau for publishing or pull out a CSV for further analysis.

It's possible I'll create one or two smaller reference tables (or just excel sheets) which I might include in the model. Tables/views would all be about 7 or 8 columns, and no more than a million rows per table.

I'd probably need to create about 4 or 5 views and bring in another 2 or 3 tables from just one database in SQL server. My question is about how well it will actually run. I've watched a few videos and the interface looks great, and I'm sure it's capable of doing what I need. So the plan would be to create a few views in SQL server to make the job easier, then connect Prep to that and do all the data cleanup, joins and pivots in prep. All the data comes from SQL Server (2016 if it matters). This will save me rewriting loads of old SQL. One of the options I'm looking at is doing all the ETL work in Tableau prep. I've taken over a project where we are moving some of our company reporting from really ancient Microsoft Access reports to Tableau.
