Tableau Reviews

Yup. It exists. And it comes in many forms. I've done a lot of Tableau presentations to managers and analysts over the past couple of years and the same concerns keep cropping up.

- The Elusive Single Version of the Truth - (the Sasquatch of data fear)

This one is usually expressed by upper management frustrated with being confused about the numbers they are being presented with and having to ask for all the details about the data in order to trust that taking action on those numbers will be worthwhile and/or safe.

Those of us in the analysts army know that this is a red herring and has absolutely nothing to do with Tableau or any other tool. There is no single version of the truth. There are 'truths within context'. A problem exists when that context isn't transparent. For example, different departments will create different versions of the same named measure with inclusions/exclusions that are pertinent to their work. So the results may be valid in one circumstance, but not when considered for another.

What is needed is proper data governance, open communication and collaboration/sharing of information. More importantly, it is the analysts duty to note definitions and sources within the reports. Remember, that report will go out into the world (or rest of the organization) without you there to explain it.

- The Dreaded Spreadmart Invasion -

This fear is often expressed by those in BI departments as it was once believed that BI departments could help us get control over the Excel invasion. Spreadmarts have existed since humans started putting numbers down on paper. It's why we invented filing cabinets. Once we all got our hands on Excel, our filing and co-ordination system became inadequate. Most organizations have a filing problem, not a reporting problem and it certainly wouldn't make sense to stop access to the analysis tool to gain control. That would be like a library ignoring the dewy decimal system and not allowing people to borrow books because it was too difficult to keep track of them.

Managing the implementation of Tableau within the organization provides the opportunity to address this problem: set up proper Projects and Groups, build and co-ordinate access to data sources, assign senior analysts to review and vet workbooks before publishing, and most importantly SHARE best practices and learnings.

- SS Data Security -

This fear is completely legitimate and ridiculous at the same time. There are already people with access to data who have no business having access to data. Try and weed them out. I have worked in places where the server (SQL) was inaccessible for days because a 'senior' analyst has left a badly built query running and forgotten about it. Seriously. My point here is that if you are concerned that certain people shouldn't have access to data connections with Tableau because they might pull all the data and freeze everything, well, they would be doing that same thing with any tool tool you gave them. They can even do it with Excel.

One of the great things about Tableau is that you can share the data if you choose. You can set up a data connection, put limits on the amount of data it pulls, do the analysis, prepare the dashboard and send the workbook to someone without having to publish it. If they have Tableau, then they can open it up and use the data to conduct more analysis. If they have rights to the connected data, then they can refresh it. You've limited the amount of data that can be pulled (e.g. rolling 12 months), so they can't crash the system. If you have Server, there's even more options.

All three of these issues can be addressed through COLLABORATION. Instead of creating rules, create conversations.

I love Tableau because they took the time to design a solution that allows me, a plain old analyst, to produce reports that help management make sound decisions on business practices. Now let me define 'plain old analyst':

- Masters in Demography.
- Advanced statistical research
- Published in peer-reviewed journals
- Courses in SPSS (grad school), Excel, Access, VBA, SQL, and SSAS (in that order).
I am not a DBA or a Business Systems Analyst. I do not work in IT. And I'm not that old.

Disclaimer: This is not intended as a braggy or defensive rant; rather I hope that by understanding who the average Tableau analyst user is, you will understand our passion for this product.

I'm an analyst because I love using data to help Directors and Managers figure out solutions to their problems. As well as building standard department/program monitoring reports, I get requests for issues that crop up as often as humans interact. Human behaviour causes outcomes, intended and unintended. Here's some examples:

- "What is the accuracy of our Cardiac Wait-list Registry in this province? People could die if our data isn't perfect."
- "How close are we to the eradication of Measles with our Immunization campaigns?"
- "What is happening with babies born with Crystal Meth addiction? Are they all coming into government care?"
- "Why are Caesarean Deliveries on the rise? What is the resourcing cost?"
- "Do we need to hire more staff or should we shift services? What's up with overtime?"
- [ENTER YOUR RECENT FIRE HERE]

Tableau saves me so much time in the creation and management of regular reports, that I can deal with these urgent ad-hoc requests in record time. Or spend more time on those requests that we knew were important but were just to complex to do in Excel without a lot of time, VBA and effort. I'm not knocking Excel; she has been a valued friend over the years, but she's a bit of a mother-in-law (she's difficult, but you still love her).

Tableau was designed with my work in mind. It doesn't limit me. It exceeds my imagination. It makes me a kick-ass user. It makes me a better analyst. So much so, that I can't STFU about it.

Tableau users are often seen as fanatics - some are concerned that perhaps we've drunk too much of the Kool-Aid. Ted Cuzzillo of datadoodle has a fantastic post on his perspective of the Analyst Users (mostly Kool-Aid drunkards) and the BI Industry Analysts who attended the 2011 Tableau Conference in Las Vegas last fall. I tend to think of Ted's BI Industry Analysts as the folks on the left side of Stephen Few's BI Wall and plain old analysts like myself on the right.

Over the years, my experience with the Tech-centric folks has been varied. Usually, I'm met with general disregard or arrogance. There is an assumption by some of these folks that us plain old analysts just make pivot tables or the occasional pretty chart. They don't consider that we may be data savy (data quality, governance, security, process) or understand that we are actual business process and analysis experts.

I work hard at trying to form good relationships with these guys, which usually begins with me having to prove my mettle or give them something they need. I try to find my "guy". He's usually the guy on the left.

He's the friendly one, the most knowledgeable one with respect to the data I need, and the one who isn't threatened by other people's expertise because he's confident in his own.

Both sides of that wall have an important role and if we work together we can learn from each other and make great things happen. I've developed some great personal friendships with people on that side of the wall... it can happen.

At the conference, I didn't meet any of the "BI experts" that Ted describes, but I did meet many other analysts and without fail, they were all nuts about Tableau and the work they were now able to do. People passionately told me about their reports, new tricks they'd learned, and the cool dashboards they'd discovered through Tableau Public bloggers.

And then there was Christian Chabot's (Tableau's CEO) opening address. Imagine a Lynyrd Skynyrd concert, the crowd is shouting "Free Bird" and the original band members are all still alive. A little over the top? Ok, maybe. But we are passionate users. Because finally, after years of wrangling and blending data in Access (or SQL if we're lucky) and spending weeks building interactive dashboards with Excel and VBA - we finally have a tool built for us.

In a post at Information Management Ted paints a not too pretty future for the tech side of BI with the adoption of new user friendly technologies, wherein senior management may not foresee the potential dangers of "receding IT". He is optimistic though; while the pendulum may swing too far in the opposite direction, a rational readjustment will likely emerge.

Hopefully it's a future where the data is well managed and prepared, and Analysts on both sides of the BI wall work together building reports and sharing knowledge. It might be a little stormy at first, but it could be the perfect storm.