Published: 07 Mar 2014
Logistic Specialties Inc. knows all too well how tangled data sets can become -- and how that can hamper business intelligence processes without the right BI and data visualization tools in place.
LSI provides supply chain management services to government agencies on defense and aerospace projects; the Layton, Utah, company also consults with other organizations to help them bid on government contracts. Several years ago, a large contract to manage the delivery of airplane landing gear to Hill Air Force Base in Utah put its analytics capabilities to the test. Answering the request for proposals and estimating the profitability of the job, and then working with parts manufacturers and tracking payments to them, were big undertakings for LSI. Data was being pulled in from numerous sources, including its own databases and government ones. When the project started, LSI's data infrastructure couldn't keep up.
"In order to get this business, you have to be fast," said Mark de Amici, LSI's chief information officer. "And when we initially did [the Air Force project], we were tracking our data on this ginormous spreadsheet, and it was just a mess."
To help clean up that mess and better meet the analytical needs of its business users, LSI in 2010 decided to implement a BI and data visualization tool from Logi Analytics. The Logi Info software lets IT teams build Web-based BI dashboards and reports with built-in visualizations through a drag-and-drop interface. In order to successfully deploy such tools and gain valuable business insights from them, it's crucial to start with a long-term plan, de Amici said.
User knowledge is good
One key element of that, he added, is to know your users; otherwise, you might end up buying BI tools and data visualization software that don't get used to their fullest -- or at all. LSI evaluated a lot of different visual analytics products, including ones from larger vendors. But de Amici said some of the tools came with "bells and whistles" that a typical user in LSI's finance or marketing department would never take advantage of. That would have left the company overpaying for functionality it didn't require. "It's feature-rich," he said. "But do you need those features? And do you have people on your staff who can really make use of that stuff?"
A lot of the buzz in the marketplace is just that. It's stuff that isn't really buying you anything.
Mark de Amici, CIO, Logistic Specialties Inc.
Another question to answer is how involved you want the IT team to be in the use of the BI system and the development of data visualizations and reports. De Amici said the goal at LSI was to enable business users to be agile in analyzing data and empower people in all departments to build their own reports, a capability that Logi supports.
SIGMA Marketing Insights, a marketing services company in Rochester, N.Y., began using Tableau Software's self-service BI and data visualization tools in 2013. Andrew Lucyszyn, a vice president who heads the company's Web analytics operations, said he wanted to empower end users to do their own analyses and visualizations because that eliminates IT bottlenecks and shortens the decision making process. "Give them [tools] where they manipulate the metrics that make sense to them, and they can produce the visualizations they need," he said.
The visual analytics market is still taking shape and generating no small amount of buzz -- and de Amici said vendors are only too happy to perpetuate the buzz if it supports their product claims. Prospective buyers need to cut through all the hype when evaluating which data visualization technologies are right for them, he advised.
For example, de Amici said he has looked at cloud-based visualization tools. But he's concerned about a lack of standards, which might make it difficult to change vendors if data can't be migrated from one cloud service to another. "A lot of the buzz in the marketplace is just that," he said. "It's stuff that isn't really buying you anything."
More to do in deploying data visualization tools
In a December 2013 webinar hosted by the Marketing Science Institute, Tom Davenport, an author and a management and IT professor at Babson College, said a number of other factors can also prevent organizations from successfully implementing and using data visualization software.
For example, project managers need to make sure that upper-level management is on board with putting data discovery and visualization tools in the hands of the average worker and that funding is in place to support an initiative over the long term, Davenport said. And failing to address data governance issues before deploying the technology can lead to big problems with inconsistent information down the road. To succeed, visualization projects require "a whole set of capabilities, from technology to organizational [ones]," he said.
Cost is another key issue, of course. Up-front costs are typically easy to understand, but de Amici said the people involved in the technology evaluation and selection process also need to weigh the long-term cost of ownership before deciding what BI and data visualization software to buy.
BI and visualization tools often are complex and require lots of development and maintenance investments over time, he said, noting that a business might have to pay consultants to help out on projects for many years to come. That model can work for large businesses but wasn't affordable for LSI, according to de Amici. "We don't have those kinds of resources," he said. "So we need something agile that isn't going to wear us out."