The Problem With Tableau Is It Always Gets Locked Down

by Plaster Group’s Data & Analytics Teamblue-door-lock-haalkab-pixabay-2

While discussing past engagements with one of my coworkers, Tableau came up. I related a story where I overheard somebody asking helpdesk for a Tableau license because her trial key expired. The helpdesk guy was apologetic and explained he couldn’t help because Tableau was “managed by a different team”. When the user asked if there was something she could use that same day, the guy said, “You can try Power BI; anybody can use that.” The frustrated user said she wanted to use Tableau. She was groaning as I walked out of hearing range.

My coworker laughed. She knew exactly what I was talking about. “The problem with Tableau is it always gets locked down,” she said. She meant that only certain people are allowed or encouraged to use Tableau, more akin to a dictatorship than a democracy.

I couldn’t help but be bothered by the irony of this statement, given Tableau’s mission statement as a company. According to Tableau’s company website, Tableau’s CEO Christian Chabot believes “Making data easy to see and understand is one of the great opportunities of our time”. If Tableau is tightly controlled and restricted, then people can’t use it to see and understand their data and the value to the organization is not fully realized. Additionally, if Tableau is locked down but other products aren’t, then users will gravitate towards those other products.

In order for Tableau to be more accessible, more used, and to deliver more value, I think IT managers and influencers of Tableau should apply the following four guidelines:


tableau-11. Create a culture that says yes more than no.

If there are plenty of licenses, say yes. As Eric Schmidt says in How Google Works, “Enough no’s, and smart creatives stop asking and start heading for the exits.” Schmidt goes on to quote University of Connecticut president Michael Hogan: “Saying yes begins things. Saying yes is how things grow.”

If there aren’t enough licenses, IT managers can create consistent and clear rules for assigning and un-assigning them. They shouldn’t create excessive processes, approvals, and meetings that passively say no (whether intentional or not): this is overhead that causes frustration and propagates bureaucracy. Instead, they should do things like maintain a visible list of who has licenses, so people can talk to each other and figure out for themselves how to share resources (which might avoid a meeting or at least email). IT managers can establish consistent end dates on used licenses, encourage users to schedule joint/paired sessions (two people one screen), and encourage people to free up licenses they don’t need (and help them get it back when they do). Whatever they do, they should try to keep the approach as similar to other products (see item 3, “don’t make Tableau too shiny” below for more on this).

tableau-22. Emphasize sharing and insights over creating something that looks cool.

Like Snow White’s stepmother, some Tableau users want to have the fairest viz in the land. They are constantly paranoid about someone else coming along and making a viz more pretty than theirs. Allowing or encouraging a “who has the best viz” culture does not necessarily help an organization see and understand its data. Instead, managers and influencers should emphasize sharing and insights, for three reasons: balancing bias, inspiring innovation, and delivering the most value to the organization as a whole.

First, sharing multiple points of view helps avoid differences in human interpretation. In his book To Save Everything Click Here, Evgeny Morozov writes, “Another critical fallacy that underpins information reductionism is its belief that an item of information can come into existence on its own, in a fully autonomous and independent fashion, without first involving an act of human interpretation.” This is why one person should not be making all the vizzes used to steer an organization’s decisions: that cool viz John made was interpreted by a human (John) and therefore carries John’s bias. Bias is missed insight.

Second, ideas from other sources inspire innovation. In Winning: The Answers, Jack Welch refers to “not invented here syndrome, or NIH”. As Welch says, “[managers] create an atmosphere where there is little interest in using ideas from outside sources to improve how things are done. […] It wrecks organizations, draining competitiveness right from their veins. […] Innovation, basically, is what we’re talking about.” There might be a better way of doing something, and innovation can find that better way.

Third, to deliver the most value to an organization, individuals and interactions should be considered in addition to the viz itself. In his book Extreme Programming Explained: Embrace Change, Kent Beck explains, “Value in software is created not just by what people know and do but also by their relationships and what they accomplish together.”


tableau-33) Tableau is a tool, and should be used.

In one organization I worked with, just mentioning the word “Tableau” caused people to look around to make sure nobody was listening and whisper, “We aren’t supposed to use Tableau.” I’ve heard people describe Tableau as if it were a sacred object that must be kept secret, taught according to ancient rituals, and used by only true believers. These beliefs are not unique to Tableau (Photoshop is another example), but are especially damaging for Tableau considering the goal to make data easy to see and understand. As an example of pragmatic use, I love this quote from Kelly Wright, Tableau’s first salesperson: “As a management consultant, I spent hours through the night iterating on charts and graphs. Tableau would have allowed me to spend more time answering questions and driving value, instead of spending hours formatting the charts.”

Managers and influencers shouldn’t treat Tableau as something that is sacred or shiny and therefore shouldn’t be touched. They should adopt a more Spartan mental model: Tableau is a tool, and should be used.


tableau-44) Help people realize how important they are to data quality and project success. Create a culture that encourages people to chase after data pipeline improvements upstream of Tableau.

Tableau is a window that can show data pipeline problems that are ugly and unhealthy for an organization. Quality now leads to increased velocity later. The cleaner the upstream data source, the cleaner the downstream viz, and the less noise for the business users making a decision. I’ve often heard people (at all levels of an organization) say, in the same detached manner as a news reporter, “The data is terrible.” Encourage your team to take at least one follow up action, and lead by example. I’ve been surprised how going one step upstream and doing one thing (one email, one question in the hallway, one ticket, whatever) can improve the data quality of multiple reports. People might not be as detached as they think, and they can make a difference.