Tableau Review
Great for users trying to perform data visualization with just Excel, but lacks control for some organizations

Pretty Pictures of BI: Tableau

There is an emerging genre of BI tools built specifically for business users to perform visual data discovery. The user interfaces are designed to be intuitive and the features simple. These software products are meant to require little if any training. Architecturally, many provide in-memory analysis for high performance.

Of course, these "exploration" tools are also not intended to be full enterprise BI platforms. Instead, they are complementary to more robust products. You will not replace your existing BI products (SAP Business Objects, IBM Cognos, IBI WebFOCUS, MicroStrategy, etc.) with the limited scope of visualization software but you might consider supplementing them.

Competitive Vendors
The main data exploration products on the market today include:

  • QlikView ($204 million revenue in 2011) 
  • Spotfire ($105M 2011) acquired by Tibco in 2007
  • Tableau ($62M 2011) 
  • ADVIZOR Solutions ($10M?) which is also sold as WebFOCUS Visual Discovery

Tibco's Spotfire is the old grand-daddy in this list and was quickly overtaken by the new kids on the block: QlikView from Sweden and Tableau from the Silicon Valley of the United States. 

While QlikTech's revenue reporting for QlikView appears much higher than that of Tableau's, the job statistics imply that Tableau is in much higher demand. 

Tableau to be the Winner?

Tableau arose out of a Stanford University research project from 1997 to 2002, and was spun off as a company in 2003. (Keep in mind that Jim Goodnight started SAS as a small college project which he later grew into a multi-billion dollar mega-software vendor.)

As part of a Department of Defense initiative, PhD candidate Chris Stolte created a "Visual Query Language" to explore large multi-dimensional databases. As luck would have it, Stolte's university mentor was Pat Hanrahan, a founding employee of Pixar. 

Together, they created what has been called “a kind of high powered, highly visual Excel,” which is a really good way of explaining the innovative software product. One of their first interested partners was Essbase, which makes sense. A common user interface for the Essbase cube was Excel, and Tableau was a nice next-generation version of that. 

Tableau is the type of visual analytics software that Microsoft itself should have added to Excel. 

The main Tableau products include: 
  • Tableau Desktop (authoring/publishing tool)
  • Tableau Server (web hosting component)
  • Tableau Reader (web viewing tool)
Tableau Desktop is an Excel spreadsheet hopped up on visualization steroids. Based on selections, Tableau points out to the user the "best practices" for visually displaying data.

Several of about twenty-four "Show Me" options light up for the user. Simply clicking on one, such as a geographic map or a stacked bar chart, does all of the work. There is no coding and no macros, just an easy to use graphical interface.

Within an hour after downloading a Windows desktop copy, I had used one of their accompanying demo files to generate a report, pie chart, bar chart, and geographic map. 

The full version of Tableau Desktop allows access to a variety of databases and publishing capabilities while the "Personal Edition" works only with desktop flat files and spreadsheets. You can easily download copies of Tableau Desktop for one or two thousand US dollars per user license (for Personal and Professional Editions, respectively). 

Why Not Tableau?

I will address Tableau's enterprise web-based and mobile usage at a later date but for departmental desktop usage, there are few hurdles to using Tableau.

If you have "spreadsheet jockeys" trying to perform data visualization with just Excel, then Tableau is a perfect holiday gift. 

If your organization has a formal software development group anxious to control BI, however, then handing out desktop tools may not be a popular option. Controlling one version of the truth becomes harder when business users create their own BI fiefdoms.

But Business Intelligence is a strange animal and IT organizations often seem unable to control it. Few want to be the BI zookeeper.  If so, responsibility for quality BI moves over to the business.

Organizations that rely upon end-user spreadsheets for reporting and analytics will bring in Tableau Desktop without much consideration.  
Disclosure: I am a real user, and this review is based on my own experience and opinions.


Anonymous avatar x30 9b84f7cf7b8e252123185ab3ccf62e1988e25e7cfbb00e1ffd48a52a5c4bc479
deepthoughtsReal User

Is there a trial version available to test and see the features? If yes, please let me know the link to download it.

Like(0)28 January 13
Anonymous avatar x30 9b84f7cf7b8e252123185ab3ccf62e1988e25e7cfbb00e1ffd48a52a5c4bc479

You can download a trial from here:

Like(0)01 February 13
Picture 1288 1365323206

You stated that its too bad Microsoft didnt include something like Tableau with Excel but it actually did as of Excel 2013! It's called PowerView, coming out I think about the time you wrote your review.

Would be very interested in a comparison between the two. I've done some primitive comparisons between Tableau, QlikView and PowerView. Seems like all of them have pros/cons. Tableau won in my mind but PowerView is brand new, its integrated with Sharepoint (if you're running SSRS 2012 in sharepoint) and is of course free for the Excel version.

One thing no one seems to talk about is collaboration. Do you really want your president making key business decisions based on some one-off PC program created by one analyst with no validation or change control? That's a disaster waiting to happen. I would argue that fancy interfaces should not be our focus but rather the quality of the data being reported. User friendliness is important but it should never ever trump accuracy. For personal data exploration, who cares what your client uses. But when it comes to enterprise reporting/decision making, you better make darn sure it's been validated and can be easily maintained, updated and shared. That means you need a collaboration platform and the ability for change and version control. Then there's the usability in terms of and integrated environment, does it integrate with your corporate website platform such as Sharepoint or is it yet another stand alone server? Or can it be standalone but supports web service calls from Sharepoint web parts, thus gaining the best of both worlds? If you're tempted to buy a tool because it has a slick interface and you've forgotten that accuracy is WAY more important than the tool's bling, that's the time to walk away because its called impulse buying. That's ok for shopping at Costco but when it comes to spending big cash for your company we owe it to our clients to make sure it's a wise and thought-out purchase that considers the real defined requirements of your specific clients.

Like(0)18 September 14
Anonymous avatar x30 9b84f7cf7b8e252123185ab3ccf62e1988e25e7cfbb00e1ffd48a52a5c4bc479
Helen WangReal User

Agree with GaryM. These few top players like Tableau, QlikView no doubt that have much fancier interface and user friendly design. I've also done a little comparison between T and Q, found it's hard to validate the data. Besides that, adding one more layer of IT resource to integrate and management control.

Like(0)18 September 14
Add aazaria

All of the tools mentioned here (you might want to exclude qlik) are front end visualizations and as such required a lot of IT involvements in the back-end.

When getting into a BI/visualization project, one must remember the data warehouse behind the scenes, else its only Excel 2.0.

Disclaimer; I work SiSense; End to End BI/Big Data Platform for business users. (

Like(0)22 September 14
Damien keogh li?1414339021
Damien KeoghConsultant

Couple of corrections if I may:
1) Tableau Desktop doesn't come in personal and enterprise versions. You can get a free trial version of the full product for 21 days but to use it for longer you'll have to buy it.
2) Tableau Reader is a free desktop viewing client, not web-related in any way. The idea here is that you can produce a workbook in Tableau Desktop and share your visualisations with others who don't have a Tableau license. This would be suitable for ad-hoc use amongst a limited population. For larger populations or higher volumes of workbooks Tableau Server is what you need.

Beyond those points, I'll add that there are many organisations using Tableau at an enterprise level. It doesn't pretend to be an ETL tool that will guarantee quality or even structure in your data, it is a visualisation tool. Often the way this breaks down is with the I.T. department working to present the data in usable format, perhaps in a data warehouse, while their business partners produce a range of enterprise reporting, insights and ad-hoc visualisations using those reliable sources. they need to work together, that's the key.

There are all flavours in between and of course some governance strucutres are necessary to avoid rogue reporting of invalid findings based on questionnable data. Tableau enables this governance pretty well. You can easily create a controlled environment where only fully verified artefacts are recognised as such, while allowing for delivery of ad-hoc explorations and visualisations of your data outside that recognised production environment. Definitely a lot easier than surrendering to spreadsheet chaos.

Like(0)22 September 14
Picture 1288 1365323206

All good points, but I might disagree with Damien on one point - that sometimes tools such as Tableau are sold as supposed replacements for ETL tools and even the need for a DW. Executives are often sold on the emotional impression that visualization tools with in-memory engines replace the need for a DW. As long as the vendor sells the tool for what it does well (and Tableau is one of the best at visualization and data exploration) I don't have problems with them. But when I have to fight the argument that these cool tools somehow replace common sense physics of data and data integration, that's when the my architecture robot starts yelling "danger Will Robinson danger"...

Like(0)23 September 14
1add tb

Gary is right. Tableau has the narrowest scope of all the next-gen highly-visual interactive publishing tools. But when an organisation is moving from printing 'one-version-of-the truth' database tables to highly-visual interactive output, this actually vastly increases the transformational work required on the data in order for it to filter and sort visually. Analysts Ventana call this new extra phase of data transformation after typical ETL 'data optimisation'. Bringing in tools like Tableau shift all this workload back upstream into the pre-existing 'SQL-bottleneck', i.e. the overworked DBAs and others who have the knowledge and skills to work directly with the database or command line ETL tools that business users will not use.

The high ROI solution comes from using a next-gen tool broader than Tableau which incorporates an end-user accessible ETL workspace and recurring data flow paradigm that allows the business users/story-tellers to make local transformations to any and all data, which are recurringly applied and refreshed automatically.

This is where the organisation's person hours are really being spent. A broader tool than Tableau uncouples the source datawarehouse(s) from the report flows, allowing each to follow its own timetable, but the reports are always on time and correct, down to little spelling changes, and auditable since all changes are fully specified in the end-user tool file-based data flow.

Omniscope is the best example of an alternative, broader end user tools that not does the interactive visualisation side side same as Tableau, but also enables end users to do the upstream transformational workspace automation where the ROI is potentially very high.

Data is not born worthy of visualisation (nor for that matter warehousing) and there is no ROI from eye-candy really. Broader, next-gen tools like Omniscope that can shift potentially all repetitive ETL/data optimisation to business users to drive (personalised) story telling should always be on the shortlist, along with Tableau. If I were IT, I know which one I would prefer.

Like(0)23 September 14
Damien keogh li?1414339021
Damien KeoghConsultant

I see where you are coming from and agree with it for some use cases, not for others.

Tableau, and other data discovery tools, have a metadata layer and calculation functionality that enable quite a lot of transformation, joining, summarising and in Tableau's case "blending" of data from disparate data sources. You could use Tableau for some ETL (or a part of ETL called "data optimisation" if you like to call it that) to produce datasets for analysis in other tools but that's not what it is for and it would be cumbersome. Tableau themselves have never sold it on this basis. There are limits to how much of this can be done and indeed whether it is wise to do it in the presentation toolset. Without tight management this approach can result in islands of repeated ETL logic across multiple applications.

There are pros and cons to each approach and an organisations needs to assess them and guage their own strengths and weaknesses. Some will decide that an IT-based ETL function is necessary to produce a standardised data layer (be it in a DW or a less structured solution) that equally enables enterprise reporting and discovery. Some will get bogged down with under-resourcing in one area (the overworked DBA!), others will find other solutions in the middle here with SQL or noSQL applications people to construct the metadata layer.

And some as you rightly say will choose a toolset that enables the business power-user to do it for themselves. We can only hope that they understand their choice and put standards and governance in place appropriate to the approach and toolset that they choose.

My own experience is that the all-singing all-dancing toolset brings with it a need for all-singing all-dancing data analysts. You fly along until the really smart guy that made it work moves on. There's lots of smart people out there but not all that many that can do everything you need. Tough challenge. If you have any scale you can normally afford to break the process down into more manageable components and choose the best tool for each one.

No hard and fast rules though, use Tableau or Omniscope for ETL if it suits you. Just be aware of what you're buying into!

Like(0)23 September 14
Jonathan friesen li?1414332822
Jonathan FriesenReal UserTOP 10POPULAR

I think the experience of my division represents where Tableau's sweet spot is. Our team is a group of business power users who help create insights into the business out of our data. Our company is a BOBJ shop but being a small division, we didn't have the funds to pony up to IT to get our own metadata layer created, ETL built, etc. Plus, IT in our organization is not embedded enough in the business to understand business req's, workflows, etc. Even in the cases where the business launches a project to create a new set of reporting/analytics, that winds up being the set of reports/analytics that users have to live with for a long time as there is a typical month+ waiting time for changes to these.

Doing some guerilla BI, our team had cobbled together access to different data sources but needed a tool to combine the data and then make a meaningful presentation of that data. When I joined the team, we were using MS Access for ETL, Excel for chart creation and then PPT for presentation. Yikes. But I give credit to the folks who did that work--very resourceful. I helped convince my manager to get a Desktop license of Tableau and it's been awesome for data exploration and presentation. It allows for a very rapid prototyping and I can update reports for our clients without a months of wait time. However, Tableau definitely could NOT do the type of ETL that we needed. We wound up getting access to a SAS server that was recently installed and SAS Enterprise Guide has become our ETL tool. It's a bit kludgy and probably isn't the optimum usage of SAS but it still gets what we need: a server-based ETL tool.

Of course now that we've established our own little rogue business warehouse, and shown how easily Tableau can present that data, IT is coming to our doorstep, as they're realizing the need to have a better level of governance over the data. We've also started building momentum with Tableau as the number of licenses quickly jumped as other users saw it in action.

For us, the value of Tableau over other tools allowed us to show what could be done, without having to wait an inordinate amount of time or spend piles of cash to get something in place.

Like(0)25 September 14
Why do you like it?