Summit Preview: Graduate to Strategy- and Process-Driven BI
Analyst-driven BI silos are no longer enough. That's the key message at this week's Gartner Business Intelligence Summit in Chicago. Analyst Bill Hostmann details trends and best practices that are putting information and insight to work.
Business intelligence is moving out of conventional, analyst-driven deployments and into the business mainstream via strategy-driven and process-driven deployments. That's the key trend that will take center stage at this week's Gartner Business Intelligence Summit, says Bill Hostmann, research vice president and conference chair. Sharing advice and using case-study examples, he offers a mini summit on the event's most important messages.
What are the trends and themes you're focusing on at this year's BI Summit?
The big message is going to be around tying together information, decision making and performance management. How do we get at the right information to make the right decisions that will drive the performance that the business demands? That's a shift from the traditional business intelligence focus, which was about query tools, reporting tools, OLAP and data-mining functionality sitting on top of a data warehouse or data mart used by analysts.
BI is segmenting into three types of applications. The first is conventional analyst-driven business intelligence. The second is strategy-driven business intelligence, in which you measure actual performance versus plans within a framework of performance management metrics. Finally, there's process-driven business intelligence, where the BI application is actually embedded within the process itself. In this case the term "application" is about connecting the BI platform at the data level, the event level and the workflow level so it has a really good understanding of the context of business decisions and the rules that apply to that decision.
All three of these categories fit together as part of a continuum. They can't be silos of analysis. They typically share information, infrastructure, definitions, dimensions, business rules or governance policies. We're trying to get organizations to look at these as part of an overall applications portfolio and BI architecture. Companies have ERP and CRM strategies, but they don't really have BI strategies or portfolios. That's just starting to emerge as a requirement.
Strategy- and process-driven deployments would naturally bring business intelligence beyond just "the BI experts." Is this changing who's attending the Summit?
We've started to get away from the term "BI" because we want to get the word "business" back into business intelligence. BI got pigeonholed as an ad-hoc query tool used by the IT organization. We want to reestablish the importance of getting information for better insight into business decisions, and we want to get the business user engaged in the discussion. We have to explain how they can use information to make better decisions, and how they work with the IT folks to bring that together.
What about the architects, CIOs and top executives? Aren't they as important as business users and IT in spreading the word?
Absolutely. The folks who manage the networks and the servers and the security are necessary, but not sufficient. Organizations have to make business intelligence a core competency. They have to know how to use information to make a decision and they have to shift the focus from managing transactions to enabling collaboration and communication. That involves an active dialogue with the business users around the kinds of decisions that they're trying to make to meet business objectives.
That a good segue to the keynote speech on "Silos, Politics and Turf Wars." What's the key message?
That's exactly why I picked [keynote speaker] Pat Lencioni. There are actually two levels of silos and turf wars that occur. One is between the IT organization and business. IT says, "Look, we built this data warehouse and you're not using it, so that's your fault." And the business users say, "What I want is more unfettered access to information, but you're limiting me to this very small subset, and it takes a long time for you to adapt to the kinds of investigation I need to do."
The other types of silos are at the business-unit level. Business unit X may have its users and IT people working in close collaboration, with a shared understanding of business rules and business dynamics and vocabulary, but then business unit Y may have an entirely different information, analysis and ways of making decisions. When you get into a larger management discussion, there's no way to compare and contrast the information from those two business units.
Is this typical in merger-and-acquisition scenarios?
Actually, organizations that are experienced with mergers and acquisitions are pretty good at recognizing those situations quickly and deciding on one approach or the other. I use the example of a large semiconductor company we worked with that had developed very independent, very entrepreneurial business units, and each had its own way of looking at customers, markets and the business. The only thing they had in common was the financial information.
The COO recognized that there had been a dramatic change in the business and they needed to restructure, but they couldn't come up with consistent baseline measures and operational insights. When the COO moved to standardize on infrastructure, business rules and measures, it led to incredible angst and anger within the organization. The business units felt like they were losing their autonomy because they had always had their own information and they never had to share it before. Once they recognized that they were dealing with a political issue, not a technology issue, it helped them frame the discussion and establish companywide standards. Eventually they were able to reorganize, and it gave them a much more agile infrastructure for supporting decisions.
Can you offer key advice from the Summit's "Organization and Best Practices" track, particularly with the new strategy- and process-oriented breeds of BI in mind?
I'll use a case study we recently published on Absa Bank, in South Africa. The bank had expanded very quickly through acquisitions, but they failed to establish a common set of performance measures aimed at meeting the business objectives of the overall bank (see "Get With the Plan" and "Q&A With David Norton" on aligning operational goals to corporate strategy).
Absa established a business intelligence competency center to define the framework of measures, and experts now go out and help each business unit establish its own tuned metrics. They created a sort of McDonalds' of business intelligence that enabled the bank to franchise the performance management infrastructure, competencies and frameworks at each of the business units. That, in turn, gave the bank a set of standards and an aggregate view of performance relative to strategy. In the process of implementing the frameworks, people better understood the strategy, so they both communicated the strategy while also helping people to understand how to tune their performance measures.
What are some best practices for process-driven deployments?
You really have to understand your business processes and their flow, and you have to focus on just a few key events or process signatures that are leading indicators of trouble or a change in conditions. Then you have to do the analysis extremely rapidly to make it effective, and you have to know what to do about it when you spot a particular event or process signature.
Many criticize BI as a rear-view mirror. So how do you do the analysis quickly and put the information to work?
If the information you develop with BI isn't actionable, it's not useful. You need to have some sort of group that lets you interact with business units so you can ask, "What does this information mean?" There are people who are really competent at both understanding the information and how it might affect the business. A group we call a "competency center" can collect these skills and insights and help people collaborate on what to do with the information you can generate. You may find that it's not good information and you should stop using it. You really have to collaborate with experts who can help you determine how to make decisions using information.
Before we close, I have to ask you about the direction of the Gartner Business Intelligence practice. Last year you lost three prominent members of the team, including Howard Dresner and Frank Buytendijk, who joined Hyperion, and Lee Geishecker, who joined AMR. Has there been a BI brain drain?
We had a few very visible folks who were at the vanguard as the market started to merge and emerge, but we've reformed the research community to be much more cross-discipline. It's co-lead by myself and Nigel Rayner, who's over in the business applications group, and we have active participation from people in the service provider community, from the vertical practices, such as Kimberly Harris-Ferrante in insurance, Hung LeHong [in retail and BI], Jim Holincheck from HR analytics and Gareth Herschel from the CRM group. We're democratizing the BI research rather than having a small concentration of a few highly visible analysts. I think that's healthy because as it becomes pervasive, it has to be a part of all these other applications and business functions.
The Agile ArchiveWhen it comes to managing data, donít look at backup and archiving systems as burdens and cost centers. A well-designed archive can enhance data protection and restores, ease search and e-discovery efforts, and save money by intelligently moving data from expensive primary storage systems.
2014 Analytics, BI, and Information Management SurveyITís tried for years to simplify data analytics and business intelligence efforts. Have visual analysis tools and Hadoop and NoSQL databases helped? Respondents to our 2014 InformationWeek Analytics, Business Intelligence, and Information Management Survey have a mixed outlook.
InformationWeek Tech Digest, Nov. 10, 2014Just 30% of respondents to our new survey say their companies are very or extremely effective at identifying critical data and analyzing it to make decisions, down from 42% in 2013. What gives?