Heavy Lifting

How Today's Economy Affects Current and Future Trends

Information Management Magazine, Sept/Oct 2009

Patrick McCollum, Craig Izydor

When we first described the major business intelligence, process and integration trends in 2007, businesses were flush with capital, CIOs were focused on strategic enterprise initiatives, business and IT leadership were aligning initiatives, and the establishment of BI and integration competency centers was all the rage. No one could have predicted the global recession's impact on BI, process and integration trends over the past months. Therefore, as we review and update the current state of BI, process and integration, we will focus on near-term trends because this recession is unlike any others we have experienced.

Since 2007, many organizations have had varying degrees of success in creating a BI foundation and aligning BI investments to corporate and IT performance management objectives. These accomplishments demonstrated the value of BI to improve enterprise visibility, drive business performance and enhance responsiveness to market conditions. We believe this has enabled businesses to more quickly and decisively adapt and react to global market changes.

Key Trends in BI

Overall, IT is maturing in establishing core and fundamental BI capabilities. Positive BI trends point to this as necessary to be successful. The IT organization will continue to:

• Enhance BI project management processes to reduce risk and budget overruns,
• Consolidate the BI application footprint to reduce application, data and support costs, and
• Closely partner with business leadership to define and justify BI projects.
• Given the economy, there are some fundamental trends regardless of industry.
• Funding is limited. Businesses will approach IT investments differently as funding is tied to company cash flow as well as market conditions.
• Capital investments are difficult to obtain and, in many cases, require approval from the board of directors. Defining compelling business value is a must.
• Investment funding is tied to stock or market performance so businesses can better manage cash flow. Deferring, shelving or canceling projects is common.
• Projects have been broken into smaller, incremental efforts that are funded within fiscal periods or quarters to closely manage capital and validate business value.
• "Quick hits" are gaining emphasis as business teams are looking to get information sooner and may not wish IT to run everything through a data warehouse for key business insights. A challenge is that they may lose attention on getting solutions backfilled in the proper architecture at a later time.

The dominating BI trend for 2009 is that CFOs want to improve financial planning and analytics. An online survey by the BPM Forum reveals that 33 percent of companies replanned, reforecast or created what-if scenarios three or more times last quarter, which is more than double the percentage (16 percent) who reported doing so in last quarter's poll. Seventy-one percent of respondents expect to replan, reforecast and create what-if analyses more frequently going forward, versus 58 percent last quarter. For midsized and larger companies it climbs to 85 percent, versus 52 percent last quarter.

BI vendors are responding to this trend in two ways. The first is bundling of BI tools and capabilities as part of business upgrades to enterprise resource planning and supply chain management systems to improve cost control and business visibility. The second trend is the creation of FP&A packaged solutions to help clients cost-effectively and more quickly leverage their investments. BI vendors that are unable to provide this capability will face a competitive disadvantage in a consolidating marketplace.

There are continued internal challenges and complexities with BI implementations. Many businesses that have invested in the BI journey over the past two years have uncovered internal challenges and complexities associated with enterprise-level BI. Many of these same issues were discovered as part of enterprise application integration activities - the need for enterprise data governance, data architecture and data management. CIOs and their enterprise architects have made positive progress in collaborating with their business counterparts. Hence, they have baseline hard numbers on the business cost and inefficiencies related to poor data management and ownership. Although common issues are identified in the BI journey and EAI activities, the challenge is how to solve urgent needs now without trying to boil the ocean and stall momentum against enterprise efforts.

Large vendors are acquiring tools and platforms that offer the client a one-stop shop for BI. For example, the acquisition of Business Objects enables SAP to offer Business Objects as an integration and reporting platform. Oracle's continued acquisitions, such as Hyperion, Siebel and Sunopsis do no less. Microsoft also continues to offer a lower-cost, one-stop option. The challenge for these firms is to meld the best of the acquisitions and carefully articulate an adoptable product integration roadmap.

Integration Trends

Data integration, the binding glue of enterprise information architecture, is making strides into merging with other aspects of the enterprise architecture. As the demand for actionable information is melded with operational data to support business decisions, it is becoming more common that data latency is driving integration efforts to respond to real-time needs. In addition, the need for shared, consistent and high quality data is the driving need for data governance. Hence, master data management, extract, transform and load/data integration tools and the disciplines that support common data integration are moving closer to critical operations. As expected, organizations still face challenges when dealing with data integration.

There is often a lack of a meaningful and leveragable enterprise data model. A predominant need exists to architect the information data model in a fashion that is consistent, easy to use and fully exploitable for multiple enterprise purposes - whether to support an integrated data hub, operational data store or enterprise data warehouse. An adaptation of the various modeling styles can produce a model suitable for the broader purpose. For example, it is generally assumed that star schema models are well-suited for data marts, third normal form models are for enterprise warehouses, and deeper normalized models support ODSs and data hubs. However, a hybrid model, which is a mix of normalization techniques, may be perfectly suitable for a data hub, EDW or ODS. This purposed model tags entities as to their meaning and usage and its design is supported by a consistent methodology. A notable style is the data vault, a term coined by Dan Linstedt, or another form called anchoring modeling. These styles foster a consistent blueprint of maintaining a model, which naturally supports the requirement of uniform integration, ease of use and extensibility. This type of information architecture style can support the service-oriented architecture and BI process management integration needs.

Page 1 of 3.