Data Analyst: Does Everybody Need To Be One?

10 Cloud Analytics & BI Platforms For Business

(Click image for larger view and slideshow.)

Will the new "freemium" Power BI, introduced by Microsoft this week, single-handedly democratize data analysis? I doubt it, but it will accelerate change in an already fast-evolving analytics and business intelligence market.

The point here is that free and easy-to-use tools are great, but be careful not to confuse widespread access to tools with reaching correct and valuable conclusions.

The stated goal of offering a free service level on cloud-based Power BI is to "open up the floodgates to all business users," according to Microsoft. The new Power BI also offers a free data-visualization tool called Power BI Designer. But free is actually nothing new in the world of business intelligence.

"All the large BI vendors have been introducing free versions of their product," said Francois Ajenstat, vice president of product management at Tableau Software, in an interview with InformationWeek. The examples date back many years.

[ Want more on analytics and BI? Read 5 Analytics, BI, Data Management Trends For 2015. ]

The entire Microsoft BI platform, for one, is the combination of free services (Analysis Services, Reporting Services, and Integration Services) included with Microsoft SQL Server together with SharePoint and free plug-ins -- PowerPivot and Power View -- for Microsoft Excel. MicroStrategy has introduced free introductory versions of its Visual Insight, Mobile, and cloud offerings. SAP has free ways to use Lumira data-visualization software. And IBM offered free access to Cognos Insight before introducing the freemium-model IBM Watson Analytics service last year.

Tableau's free option is Tableau Public. Ajenstat said that more than 70,000 users have registered, downloaded the vendor's desktop software, and posted data visualizations on the Tableau Public site -- generating more than 300 million views in the process. But Tableau is about more than visualization, he insisted.

"We help people answer deeper questions from their data by making it easier for them to interact with their data and ask new questions," he said. "You discover the real meanings in data by testing hypotheses and chasing hunches. It's not just about showing an end result; you have to show how you got there and defend your answer."

This point gets to the difference between having a tool and drawing correct conclusions from your data. Ajenstat's it's-about-the-process view trusts that people are intelligent, have a basic understanding of their data, and will explore and rigorously test their hypotheses.

A Tableau Public visualization of New York City graffiti data.

There's also a possibility that people will bring biases to an analysis, that they may not consider all relevant possibilities, and that they may not have enough information. Say a company wants to get to the bottom of a shortfall in sales. A marketing leader might look at the data and argue for a larger marketing budget, noting the sparse pipeline. A sales leader might conclude that more salespeople are needed, seeing low engagement and close stats compared to available opportunities.

"Both of those could be great insights related to the problem, but they may not be the cause of the problem," said Marcus Hearne, a business unit executive leader of IBM Watson Analytics, in an interview with InformationWeek. "It could be that a competitor suddenly cut prices by 50%, causing customers to not really take any interest in the company's products."

Making the analysis intelligent is just as important as making analytics accessible and easy to use, said Hearne. IBM Watson Analytics does this, he said, first by surfacing many possible interpretations of data in a ribbon of

1 of 2