What is most valuable?
It's easy to start using to deliver off-the-shelf visualizations. It makes it easy to share dashboards with a team. It's a comprehensive solution that can scale from desktop (analyst, developer) to server (large end-user audience) with small or no changes.
How has it helped my organization?
In my experience, its greatest achievement has been opening stakeholders' minds. It pioneered the self-serve trend (which may have social roots) where people "own" their data and IT is a service provider to ensure visualization content is based on robust data. Connecting to a massive number of sources (that can be blended in Tableau) has also contributed to this achievement.
What needs improvement?
It came at a cost. Developers end-up, many times, scratching their left ear with the right hand, in order to make visualizations more powerful, effective or appealing than off-the-shelf ones. The next challenge for Tableau will be combining ease of use for basic-to-mid level complexity design (self-serve, "power users") with flexibility and repeatable development in a corporate environment (professional designers), avoiding "hat tricks" or "cookbook" style.
Also, when building corporate dashboards, it may be challenging to gain the type of control needed without having to resort into some forum tricks or Zen-Master's advice. Workbook, connection versioning, and governance would be great for mid-to-large size organizations. Server-wise, having options to trigger alarms to an Administrator under certain conditions would also be an asset.
For how long have I used the solution?
I've been using both Tableau Desktop and Server for four years.
What was my experience with deployment of the solution?
No problems at all. Very robust and well documented (both Desktop and Server versions)
What do I think about the stability of the solution?
No problems, at least with the data sources I have used Tableau so far - MSSQL 2008 & 2012, Teradata RDBMS 14.x, and MSAS Cubes.
What do I think about the scalability of the solution?
Tableau can scale to massive datasets without problems (I used a few, approximately one billion-row tables), provided a DBA is part of the team to fine-tune tables and views. The proprietary columnar database engine is very fast and robust.
How are customer service and technical support?
It's great. Technical Support
It's great, although forums are usually the first resource to get answers from.
Which solution did I use previously and why did I switch?
I have helped customers migrate from MS Pro-Clarity and Oracle Hyperion.
How was the initial setup?
Tableau is not a demanding platform in pre-set environments before rolling-out as long as tech specs are met. However, certain design decisions need to be made in advance to ensure good performance and smooth scalability on a Server deployment (e.g. permissions, standards, etc.).
What about the implementation team?
I act as an implementation consultant.
What's my experience with pricing, setup cost, and licensing?
Go through the exercise of planning short and mid-term to understand what the best licensing model is e.g. per Core vs. Named Seats, number of Designers, etc.
What other advice do I have?
Understand your organization's profile. Do you have or want/need to develop power-users or will you use traditional IT developed dashboards in an intranet/internet? Although Tableau can fit both scenarios, the first option would be the most benefited one when compared to other solutions. Look into redefining the relationship between business stakeholders and IT, since the former will demand more consumable information for analysis instead of packaged reports.