Monday, September 21, 2009

Want to make BI pervasive? It's the culture, stupid


Business intelligence software may have been around for several decades, but it remains an esoteric niche in most companies, according to an analyst.
Unfriendly corporate cultures, not the BI tools or apps themselves, are preventing BI from becoming pervasive.
"The technology has been around for a long time. It's the people that often get in the way," said Dan Vessett, an analyst with IDC Corp.
IDC recently conducted a study of 1,100 organizations in 11 countries measuring how pervasive BI is in companies, what factors helped make it more pervasive, and what "triggers" data warehousing architects and IT managers can use to the further the spread of BI in their companies.
In a speech at a conference in Chicago, Vessett said IDC measured BI's pervasiveness via six factors:
  • Degree of internal use. According to IDC, that was between 48% to 50%.
  • Degree of external use, or how much it shared data with vendors or customers. Sharing BI data keeps customers loyal, Vesset said. And canny BI users in industries such as retail can sell that data to generate non-trivial revenue, he said.
  • Percentage of power users in a company. The mean was 20% in surveyed companies.
  • Number of domains, or subject areas, inside the data warehouse. Over five years, the average at surveyed companies grew to 28 from 11.
  • Data update frequency. While real-time updates can be indicative of heavy dependence upon BI, "right-time data updates" is more important. "Daily, weekly or monthly could be sufficient," he said.
  • Analytical orientation, or how much the BI crunching helped large groups or the entire organization make decisions, rather than isolated individuals. "The fact is that most individuals and companies are not data driven. They still rely more on experience rather than analytics," Vesset said.
According to Vesset, these factors in descending order had the most impact on BI pervasiveness:
  • Degree of training, not in the BI tools -- "the vendors do a pretty good job" -- but in the meaning of the data, what the key performance indicators (KPIs) mean, etc.
  • Design quality,or the extent to which IT-deployed performance dashboards are able to satisfy user needs. Satisfied users will talk up the BI software, creating "BI envy" in other employees, helping spread the software's use. Unsatisfied users will go around IT and use Excel or some SaaS applications. Prominence of the data governance group.
  • Involvement of non-executive employees.
  • Prominence of a performance management methodology.

Vesset also listed a number of potential "triggers" for BI projects that IT should take advantage of:
  • Arrival of new executives, who, if unsatisfied with the type of reports or analyses delivered, may help sponsor a new project.
  • Need to comply with new legislation.
  • Introduction of performance management methodology.
  • Corporate reorganizations, including mergers and acquisitions.
  • Changes in the organization's growth, such as when a fast-growing company slows down and then begins focusing on improving its profit margins.

Saturday, September 5, 2009

Smarter Buying: Business Intelligence and Performance Management Software


Here's what matters when selecting a product and supplier. Hint: Price isn't everything.

Boil down the scores of industry polls of CXOs over the last few years, and you get a remarkably consistent message: We need to become a smarter, faster, more efficient (and cheaper) organization. No surprise this now-familiar mandate has been adopted by numerous industry vendors.

Nowhere is this truer than the hot markets for business intelligence and performance management software and services. IDC forecasts global sales of BI software tools will grow from $7.5 billion in 2008 to $10.2 billion by 2013, or about 6.3 percent annually.

It makes sense: Who doesn’t want to make better organizational decisions? Run a tighter, more effective outfit? (Especially, as vendors are eager to point out, during an economic downturn.)

Finding the most appropriate decision-making software will take some thoughtful decision-making of your own. As with much in IT (and life), the answer to “what’s best?” depends heavily on your particular situation.

  • What strategic vendors have you committed to?
  • What’s your budget?
  • Staff expertise and availability?
  • Are you looking for an easy-to-use solution for a group of nontechnical business users? A master analytic engine for the entire enterprise?
  • Are you willing to try a lesser known small but maybe more affordable supplier?

Very basic stuff, but here as elsewhere, they are make or break considerations.

That said, the list below shows what your IT peers look for when buying BI/PM software and services. It’s from an e-mail and online survey of 1,380 qualified buyers of products in the space conducted for Ziff Davis Enterprise by Preference Research in January.

Respondents were asked whom they now buy from, and whom they’d consider and recommend. Predictably, the list of BI suppliers that respondents would strongly/very strongly consider for next purchase was dominated by the Big Five in the space: Microsoft, Oracle, Business Objects, Cognos (IBM) and SAP. Predictably too, each vendor had particular strengths: IBM had best combo of features and reputation, Microsoft had best user familiarity, Oracle and SAP scalability.

On price, respondents all wanted cheaper products (stop the presses!). Fortunately, they also said price mattered less (ranked seventh overall) than good ROI (third).

Here’s what buyers told us what matters in choosing BI/PM. How does that square with your checklist?

Attributes "Very Important" to Selecting Vendor for Short List:

  • Usability: e.g., intuitive interface: 71.8%
  • Performance: e.g., speed, stability, lower rates: 63.1%
  • Value for the dollar: i.e., good return on investment: 50.6%
  • Features: e.g., uniqueness, depth, superiority: 49.6%
  • Technical support/service: e.g., telephone, Web, on-site assistance: 49.3%
  • Scalability: e.g., ability to handle high-volume usage: 44.7%
  • Price: i.e., relatively inexpensive: 41.8%
  • Low cost of operation: e.g., ease of deployment/maintenance: 40.0%
  • Familiarity: e.g., convenient because you or others already use it: 18.9%
  • Reputation: i.e., positive word-of-mouth, news or reviews: 18.7%

Operationalizing Business Intelligence


Linking business intelligence with business results gives workers out in the field better tools to drive day-to-day operations and customers better ways to make informed purchases.

At the advent of business intelligence, the idea was to put the right data and analytics in the hands of people who could make actionable changes that improve the way business is done. Somewhere along the line, that simple idea grew muddled.

BI systems grew up to be scattered across enterprises with the wind, complicated and difficult to use by even the business analysts. As enterprises assess how to move forward with their BI efforts, one of the driving forces of these initiatives is to make BI simpler, easier to access by a wide range of workers. In short, organizations want to bring BI back to its philosophical roots.

“One of the promises of BI when I started was empowering decision-makers and knowledge workers. It was to create pervasive BI and leverage BI for everyone,” says Dyke Hensen, chief marketing officer for PivotLink, who calls himself an old BI "oak tree" after 20 years in the space. “The problem is that over the years a lot of these offerings became very complex, very bloated and expensive.”

Hensen cites figures from The Data Warehousing Institute annual survey that showed the median cost just to maintain BI applications alone clocks in at around $235,000 per year. In his company’s case, Hensen says the goal is to reduce the cost of maintenance by offering BI capabilities via a SAAS (software-as-a-service) model to reduce not just the hardware and software costs, but also the number of employees needed to maintain systems.

However, according to Nimitt Desai, business intelligence and data warehousing lead for Deloitte Consulting, many organizations can’t even feasibly begin to leverage SAAS until they begin to consolidate their BI efforts. One major problem enterprises face today is the sheer number of BI applications spread out over an organization. Desai says it is common to see enterprises running well over 100 analytic environments that they must report against.

“When you have hundreds of systems, then SAAS is a myth,” Desai says. “But if you have a smaller amount of sources, I feel there is a big push in that direction.”

This drive to consolidate sources is much more of a possibility today than even two years ago with the push by major ERP vendors to help bring BI out of the cold and under a larger operational umbrella. Acquisitions such as the SAP pickup of Business Objects last year are a sign of where the BI space is headed.

According to Wayne Eckerson, director of research and services for The Data Warehousing Institute, this shift to bring together operational systems and BI just makes sense.

“It is kind of odd that you have to switch contexts, if you're an operational worker, from an operational app in order to open a dashboard or a report to understand what the impact was of what you just did or see the context of an action,” Eckerson says. “There is definitely an opportunity for vendors to take that gap out of the BI office and embed BI right into operational applications.”

Brian Kilcourse, managing partner for RSR Research, agrees that this "operationalization" of BI is one of the most impactful intelligence trends sweeping through enterprises at the moment. He says he’s seen lots of anecdotal evidence of how a shift to better embed BI within operations gives workers out in the field better tools to drive day-to-day operations or customers better ways to make informed purchases.

“We’re seeing a lot of companies injecting actionable information into operational processes in just-in-time fashion,” Kilcourse says.

For example, in one case study Kilcourse analyzed, he witnessed Virgin Megastores offer its store managers a strong way to improve sales. BI systems there were integrated with up-to-the-minute in-store sales so that managers could see how hit titles were selling in comparison to other hits with similar sales. The intelligence match-up compared the first few days of release of one title with other releases that had similar sales starts and gave managers the ability to project outward. It also offered actionable analysis that enabled workers to pair up other overstocked albums with hot sellers in endcaps to move otherwise stationary products.

Even though Virgin closed its retail stores for other reasons entirely, Kilcourse says this application of operational BI is too good to go ignored.

“They were basically doing a kind of a product mashup on the sales floor in more or less real time based on the signals they're getting from sales as they're occurring,” he says. “So they're basically doing shelf resets based on the fact that one title is flying off the shelves and they want the other one to fly with it.”

Kilcourse says that these sorts of initiatives help organizations better adopt a sense-and-respond type of mentality. He also believes that better embedding BI into operations provides very good back-end benefits.

“One of the big values of it is that the operational systems or the processes can then deliver back some information to the business intelligence system that says, ‘This is what happened after you responded.’”