Tableau Software’s Chabot: Not Your Father’s Business Intelligence

January 5, 2009

Listen to the Podcast:

–By Christian Chabot is CEO of Tableau Software

Called decision support in previous decades, today’s BI industry was built between 1988 and 2005 by companies like Brio, Crystal Decisions, Cognos, Business Objects, Hyperion, Information Builders and MicroStrategy. For years, demand drove exceptional performance in the industry. But more recently, the BI industry has witnessed a wave of consolidation. In the last three years alone, Oracle acquired Hyperion; IBM acquired Cognos; and SAP acquired Business Objects. These companies are now entering their sunset years.

Why? Because the most important trend in business analytics today is that the conventional “BI Platform” approach promulgated by these vendors is dying. In fact, in today’s environment, BI platforms are built on opposing principles to what is needed to create analytical advantage.

A new battleground is shaping, and it’s not your father’s business intelligence.

Competing in the New Environment: Data is Everywhere

BI companies have traditionally sold business intelligence platforms or business performance management platforms that promise to take businesses out of spreadsheets and disparate databases into centralized decision making platforms. This was accomplished by creating platforms primarily designed for report distribution. In this approach, data is centralized before people can start analyzing data. Reports are developed by BI experts, usually well in advance of any business person seeing any results.

Once the system is deployed, people who need specialized reports or new data sources must file change requests with the people running the BI Platform. Since only a few programmers have been trained well enough to handle these requests, changes are often weeks in the making. While they wait, business people make do. They’ll pull the original report into Excel where they do limited and usually inadequate manipulation.

This approach is not keeping pace with the inflow of data. Data is now everywhere and growing exponentially – extracted from SAP, downloaded from web sites, flowing from applications like Salesforce, NetSuite and Google Analytics, burgeoning in spreadsheets, saved in PDFs and spilling out of SQL databases.

The BI Platform approach in this environment has become a bottleneck. It is taking too long, costing too much and delivering too little. IT departments are frustrated. Business users are frustrated.
In a typical scenario a company starts a BI project and invites some BI Platform vendors to pitch the business. The vendors convince the company that all of their business intelligence needs should be centrally managed in a “uniform and strategic approach to business intelligence.” This is their most expensive option. License and maintenance fees instantly exceed $150,000 and quickly surpass $450,000. Implementation experts are hired. Services bills mount. New headcount is opened. The sales people working for the BI Platform vendors toast their success over steak and wine. Another BI Platform bloodbath has begun.

Back at the company, months pass and the project schedule starts to slip. The main cause is that the project is overloaded with the needs of too many people. A second problem is that the BI Platforms are difficult to deploy. Designed in the 1990’s, they are cumbersome and slow. Just getting a typical basic BI server up and running can require the installation of a dozen different server components. A third problem, and potentially the most critical, is that the end-user interfaces are simply too hard to use. So even when the system is finally up and running, people do not use it.

It becomes evident to the company that the BI Platform approach wasn’t clearly thought through, and questions begin to emerge:

Did we need a BI Platform to solve a departmental analytics project?

Did we evaluate how difficult implementation would be?

Did we evaluate how difficult the end-user interfaces would be?

Did we consider how difficult change requests would be?

Did it make sense for the buyers of the platform to be different from the users?

In the next phase, the project is trimmed down so it can deliver something, however small, to some small group before the end of the year. Results are delayed. Expectations fall. Competitors surge ahead. In the new fiscal year, more budget is allocated to try to fix the problem. The “uniform and strategic approach to business intelligence” isn’t looking very strategic after all.

Competing in the New Environment: The Solution

What is needed in the new environment are rapid fire analytics technologies that can be instantly applied by business units and are extremely easy for people to use. Even better, they let business users ask their own questions. Thanks to advances in technology made this century, such BI solutions are now available. Tableau Software, as an example, now boasts thousands of customers, including Google, Apple, Wells Fargo, Cornell University, Cirque de Soleil, Verizon, Coca Cola, Jabil Circuit, Sony, Dow Jones, Pfizer, the US Air Force, Wal*Mart, McKinsey, Royal Carribean, and even PayPal.

The new generation of tools is not designed to replace BI Platforms completely. On the contrary, they are often used to take pressure off of centralized BI Platform projects, ensuring that the needs of high impact teams and groups (e.g., finance) are met while large scale reporting projects slowly advance. In short, companies that take advantage of this new generation of BI tools to go ‘beyond centralization” are building the new competitive advantage.

The primary attributes of rapid-fire BI systems are as follows:

People can be up and running in minutes.

People can access new data on-the-fly without assistance.

People can bring new data in and out of their work easily.

The applications have forgiving user interfaces with “unlimited undo.”

The applications support different meta-data models, not just a global master.

They applications can be trialed and validated with no cost.

People can self-train at no cost.

With the new generation of tools, people keep pace with the inflow of data no matter where it is or how it grows. Deployment is painless and requires little maintenance. And most importantly, people actually use the applications.

Here is a checklist to help determine if you are looking at a new generation business intelligence solution.

Top Five Signs You are Being Duped by a BI Platform Vendor

Demos: Your initial meetings with the vendor do not contain product demonstrations. Instead, there is lots of PowerPoint.
Recommendations: Their presentations contain complicated diagrams of “BI Stacks” and “Platforms.” They advocate nonsense terms like “Enterprise Wide Real-Time Decision Making.”
Engineers: They tell you their BI Platform is easy to use, but every time you ask a non-trivial question, sales engineers get involved.
Price: You can’t find prices on the vendor’s website. And when you ask a sales person, he can’t tell you until he learns more.
Quantity: You can’t start small (literally 1-5 users) and prove the results before proceeding with a large investment.

Prior to joining Tableau, Christian was an Associate Partner at Mobius Venture Capital, a $2.5 billion venture capital fund, where he specialized in enterprise applications. Before Mobius, Christian was CEO and co-founder of BeeLine Systems, a pioneer of next-generation digital route mapping technology. BeeLine was acquired by Vicinity Corporation (NASDAQ: VCNT) in 2000. Christian also brings ten years of analytical software experience from his work at Cornerstone Research and Siemens AG to Tableau. He is the author of “Understanding the Euro: The Clear and Concise Guide to the Trans-European Currency” (McGraw-Hill, 1998) - a business bestseller. He holds an M.B.A. from Stanford University, an M.Sc. from the University of Sussex, England, and a B.S. in Engineering with Honors from Stanford University.

Posted by Carl · Filed Under Interviews, jobs