In order for readers to view and open workbooks using Reader, they will need to have a packaged workbook. This means they will have access to the underlying data. Another thing to consider with this approach is the size of the workbook and data. If working with large data sets and workbooks it may prove tasking to distribute.
Have you considered Tableau Digital? It offers the ability to restrict access to underlying data.
If security is your top priority then Server would be the best option, though As digital is designed for public websites and audiences, anyone will able to see the published views.
Hi Andrew, I can't give you the pros/cons, but I can tell you what we decided to do, and how it's worked out. We chose the Desktop/Reader route and it has worked out quite well. Any 'sensitive' workbooks are placed on an FTP server, which the users can then download. None of our workbooks are of any significant size, so this isn't an issue for us.
Even our least computer-savvy clients are able to figure out how to install Reader, and then open the workbooks we provide. In most cases the data belongs to the client, so we're not concerned with exposing the underlying data. We originally thought we'd give Tableau a try and then upgrade to Server. So far we haven't had a need to do this. But we've only been using Tableau for about 6 months.
Tableau Server is fantastic way of sharing interactive Data Visualizations, Mobile Business Intelligence and gives the users the ability to share insight by advertising customized views (new in 7.0.2) within your organization. Also, Tableau Server can be used to build a enterprise security model around your data while hosting Data Visualizations, users can see only what they are allowed to see (great for financial, HR and sensitive customer data). If you are considering Server these should be towards the top of your needs list.
From a development standpoint, distribution of the end product is instantaneous and as database data sources are updated the Visualizations will requery the latest data without the need to maintain and distribute packaged workbooks. Further, with the Data Server, Tableau Extracts and Tableau Data Sources can be shared across the entire data analysis organization. This means each analyst can start with the same data sources, meta-data and calculations, They will love that.