RESEARCH & RESOURCES

Cindi Howson

From the BI Scorecard Blog: Deciphering the Buzz in Dashboards, Visualization, and In-memory

Ultimately, the question is: what are you trying to accomplish?

The BI industry abounds with buzz words, lending excitement but also confusion. So, too, with some of the hottest topics today: dashboards, visualization, and in-memory.

These market segments are closely related, sometimes overlapping and sometimes not. It’s made for some interesting evaluations and bedfellows in recent months. I puzzle at who is now brought together on customer short lists: QlikTech QlikView is compared with TIBCO Spotfire. Microsoft PowerPivot is now compared with Tableau. Customers seem unsure if they should compare SAP BusinessObjects Explorer to QlikView or Xcelsius to QlikView.

Nobody wants to waste money and time investing in competing technologies. The key to comparing and investing in the right products is in understanding their sweet spot and their differentiators. Ultimately, the question is: what are you trying to accomplish?

Dashboards: Information at a Glance

I like author and expert Stephen Few’s definition of a dashboard: “A dashboard is a visual form of information display, which is used to monitor what’s currently going on in the business at a glance.”

Wayne Eckerson expands on this definition to say, “A performance dashboard is a multilayered application built on a business intelligence and data integration infrastructure that enables organizations to measure, monitor, and manage business performance more effectively.”

A variety of technologies can be rightfully referred to as a dashboard: a spreadsheet with multiple visualizations can be the basis for a dashboard, for example. The industry has come a long way from the days of static executive information systems of the 1990s. Even today’s second generation dashboards are a vast improvement over solutions on the market just four years ago, as shown in the below table.

Better dashboards today: Versus first-generation dashboards that are:
Are user assembled Custom coded or built by IT experts only
Are highly interactive Static
Use space intelligently Overwhelmed with too much data
Smartly communicate information with appealing visualizations Use so many animations and eye-candy that it’s distracting
Are updated as needed Updated monthly or quarterly
Analyze and interact at the speed of thought Slow and have unpredictable response time

Dashboards are ideal for executives and managers who need to monitor multiple business aspects and track key performance indicators. For dashboards that are updated in near real-time, operational staff such as call center operators benefit from dashboards that allow them to monitor sales and service performance, order fill rate, or compare performance to the rest of the team. Dashboards can help your company ensure everyone is working in the same direction, monitoring achievement toward common goals.

The bottom line is that if dashboards have not been part of your BI tool strategy to date, you should make it a priority in 2011. Your dashboard solution can come from a BI platform vendor, a specialty dashboard vendor, and, in some cases, a visual discovery vendor. However, not all visual discovery vendors have dashboard capabilities, and not all dashboard products rely on in-memory capabilities.

Some examples of vendors that focus primarily on dashboards include QlikTech, Corda, iDashboards, and Dundas.

Advanced Visualization: Discovering What Matters

Advanced visualization and discovery is a well-established software segment, but it is one that is getting increased excitement in the BI community. Historically, advanced visualization was a tool reserved for a specialist community, usually to answer less well-defined questions. For example, in life sciences. an advanced visualization product might be used to identify patterns for drug side effects. Did side effects only occur in certain genders? Age groups? Life style characteristics? Unlike a business query tool, in which a user asks a specific question and gets a tabular result set, an advanced visualization tool may display all the data graphically and visually reveal patterns that a user may not have been explicitly looking for.

Beyond this difference in use case, advanced visualization tools have generated excitement because of their sweet spot in visual appeal. As I’ve written before, “a picture is worth a thousand numbers.” Bored and unenlightened by a dense screen of numbers, business users like how an advanced visualization tool can speed the time to insight and be so much more eye-catching than a traditional business query or reporting tool.

Visual discovery tools have been maturing to the extent that some of these vendors claim to be able to replace a large portion of the requirements filled by a BI platform. In some cases, they can. In others, they are more appropriately complementary to a BI platform. More mature products may have added Web-deployment capabilities, security, information consumer features, and dashboards. However, a few advanced visualization products continue to be better positioned for individual analysts.

Some examples of visual discovery products with dashboard capabilities include Tableau, TIBCO Spotfire, and Advizor Solutions. SAS JMP is an example of a visual discovery product but without dashboard capabilities. SAP BusinessObjects explorer has some visual discovery capabilities but is primarily a search and navigation tool, without dashboard capabilities.

In-Memory: Insight at the Speed of Thought

In-memory analysis is a related technical innovation that can power both dashboards and advanced visualization tools. In-memory technology has been around for decades as I wrote in this Information Week article. Two main things have changed, though, that has accelerated the adoption of in-memory analytics. The first is 64-bit operating systems that have more addressable memory, in theory up to 1 Terabyte of data. With 32-bit operating systems, tools that relied on in-memory were less scalable.

The second is the need for speed amidst every-increasing data volumes. Business survival depends on faster, smarter insights. The recession has shown multiple lessons of survival of the smartest (Best Buy versus Circuit City, Continental Airlines versus Delta Northwest Airlines, or closer to home in New Jersey and smaller: Carlos’ Bakery versus The Wyckoff Bakery).

IDC recently estimated that in 2010, we will generate 1.2 zettabytes (1 trillion Gigabytes) of new data. With so much more data to analyze, users can’t wait months (batch query) or even minutes (scheduled query) to get answers to critical questions. The discovery has to be at the speed of thought. Loading data into RAM is one way to do that. Often vendors will take multiple approaches to achieving speed of thought, in some cases leveraging the hardware through analytic appliances, a columnar database for storage optimized for analysis versus transaction process, and increasingly, storing data in-memory.

In-memory analysis is a technology to pay attention to. However, I get concerned when customers say they need to find an in-memory solution. In-memory should not be a panacea and I don’t view it as its own distinct BI segment. Instead, it’s a technology that crosses multiple segments. It is but one of several techniques a vendor may use to arrive at speed-of-thought. As a technology evaluator, your goal should be to know when a vendor uses it, how efficiently, and what else they use to achieve speed of thought analysis.

All three of these industry innovations are important to improving business intelligence. Understand their value and how they complement your existing BI tool portfolio. Although customers may wish that these segments converged, that’s not the case today. The first step in exploiting them is to understand the distinct value each provides, and then assess where to complement your existing BI tool portfolio.

To help you understand more about these segments, I will be teaching a new course at the next TDWI World Conference in Las Vegas, February 15: Evaluating BI Dashboards with Bake Off. In this course, SAP BusinessObjects, QlikTech, and Tableau Software will be performing live scripted demos. A different set of vendors will be participating in TDWI April 5, in Washington, D.C.

Best wishes,
Cindi Howson, www.BIScorecard.com

Copyright © 2011 by Cindi Howson and BI Scorecard; reprinted by permission of the author

TDWI Membership

Get immediate access to training discounts, video library, BI Teams, Skills, Budget Report, and more

Individual, Student, & Team memberships available.