Let’s face it: unless your organization has standardized on Tableau for enterprise or self-service business intelligence (BI), the tool can wreak havoc on your BI strategy and operating environment.

Like a wild vine, if left unchecked, Tableau will strangle your BI program, creating ungoverned spreadmarts that increase your support costs, undermine data consistency and waste your staff’s time reconciling reports.

Let me be clear: there’s nothing wrong with Tableau. In fact, it’s a very good analyst tool for visualizing data sets, especially spreadsheets, and it makes pretty good dashboards, too. The truth is Tableau is the new Excel—a spreadsheet on steroids if you will—that has taken the business community by storm.

With Tableau, a business user can buy and download the product and begin visualizing data sets in minutes. They don’t need to consult corporate BI or information technology (IT) teams to become productive with data—and that’s exactly the way most business unit managers and analysts prefer it.

Tableau has struck a chord with business users. But it’s not a new phenomenon. It fulfills the same need that Excel has served for the past two decades. It’s just that Tableau offers a better analyst and user experience than Excel.

Sales Strangulation

The catalyst behind Tableau’s growth is its “land and expand” sales strategy that targets business unit managers and analysts and largely circumvents BI and IT. The idea is that if you can get enough individual sales in enough departments, you can create a defacto enterprise standard that corporate BI/IT has no choice but to accept.

This kudzu-like sales strategy has been hugely successful, especially in organizations without strong BI/IT standards for tools and data, and even in some with published standards where business users have high dissatisfaction with incumbent tools, processes, and data. Tableau has successfully ridden this wave all the way to a successful 2013 $250 million initial public offering and beyond.

Unfortunately, this sales approach has ruffled the feathers of some BI directors who haven’t standardized on Tableau, but have developed an enterprise BI strategy that intelligently maps business roles to BI tools and data structures and provides ample training and support to promote adoption. Here are some comments from the BI Directors Subgroup of the BI Leadership Forum, which I moderate:

"My concern is that Tableau's entire marketing approach appears to be to sell to end-users with the expectation that this will eventually create enough demand to push IT organizations to adopt and support their product without regard to fit with the overarching data visualization, data platform, data integration strategies as well customer and employee collaboration strategies. This seems to pit business users versus IT and produce the traditional counterproductive outcomes of this approach."

In organizations where Tableau runs rampant in an ungoverned fashion, there is little beleaguered BI directors can do. At best, they can wait for Tableau users to “hit the wall” and experience data complexity or performance issues that are beyond their technical know-how to fix. At that point, BI directors can reassert governance and rules regarding the role of Tableau in the organization and how to install and architect it to avoid issues.

Enterprise or Bust?

Like any BI vendor, Tableau is eager to notch enterprise deals—after all, that’s where all the money is. And your biggest customers have an uncanny way of requesting new features and functions that a publicly-held software vendor ignores at its peril.

In fact, today’s incumbent BI vendors—Business Objects, Cognos, MicroStrategy, and Hyperion—started life as desktop tools for business analysts and gradually migrated upstream to departmental deployments and then enterprise ones, which required massive and painful remakes of their product architecture and go-to-market strategy. (The first three then soon after got acquired by enterprise software companies.) What were once lightweight and relatively inexpensive desktop tools  are now heavyweight enterprise tools that carry heft price tags.

Will Tableau Follow Suit? The question is whether Tableau will follow suit of its BI forebears or remain an agile desktop tool designed primarily for business analysts. If it wants to compete in the enterprise market, it will need to make a few changes.

For instance, it has limited ability to integrate data from multiple toolsets, does not offer a robust modeling or semantic layer, offers limited support for object reuse and templates, runs into scaling problems with large data sets, doesn’t fully support production reporting and report bursting, doesn’t have a robust server environment, and has limited analytics compared to other tools.  It’s also very expensive since Tableau rarely offers volume discounts on its $2,000 per user desktop license.

Enterprise Sales. Perhaps most importantly, it doesn’t know how to sell enterprise deals or at least avoids discussing enterprise BI deployments with BI/IT professionals. The same BI director quoted above in the BI Leadership Forum subgroup added:

"My issue with Tableau is what appears to be a lack of enterprise sales capability…. where [the vendor needs to participate in] larger scale strategy sessions, develop roadmaps and ROI proof points, and conduct pilots to land the big deal…. We are engaged with the Tableau sales team but each discussion seems very limited and then we experience the direct selling model to our end users and this creates confusion for our users and more work for us to communicate again the tools we’ve selected to date."

Riding the Tableau Wave

The point of this article is not to complain about Tableau but recommend ways to manage a rising tide of Tableau seats in your organization. The nine recommendations below are gleaned from numerous BI directors with whom I’ve discussed this issue during the past several years. I suspect you can probably add a tenth recommendation so please add your comments at the bottom of this article!

1. Symptom or Problem? It’s important first to understand why business users are adopting Tableau. If their information delivery needs are not being with existing tools and processes, then the proliferation of Tableau is a symptom of a deeper problem, not the problem itself.

2. Complementary Standards. Many companies that have standardized on Tableau to meet the needs of business analysts have also standardized on another tool for enterprise dashboards and reports. The two are not mutually exclusive.

3. Get Out in Front. Don’t wait for the business to bring in Tableau (or another visual discovery tool), introduce it yourself and deliver a few quick wins. The business will see you as a committed partner with their best interests at heart.

4. Standardize Data and Metrics. If you preempt business users and install Tableau yourself, make sure you point it at your data warehouse (or other trusted sources of data) and populate server-based Tableau workbooks with standard metrics and dimensions to ensure data consistency.

5. Establish Communities of Practice. Create an in-house forum for Tableau users so they can share best practices and tips and support each other, lessening the time your staff has to spend supporting users.

6. Work with Tableau Sales Reps. Ask Tableau sales representatives (as well as your business users) to notify you once a business user decides to try or buy the product. This keeps you in the loop so can gauge sentiment and trends.  

7. Non-Supported Tools. If Tableau is not an official company standard, explain to business users that BI/IT doesn’t have the resources to support the product and they will have to support themselves.

8. BI Portfolio. Explain the rationale to business users of your existing BI tool standards. Communicate that you want to provide them with the right tool for the task. Usually, this engenders a greater degree of trust.