What’s Hot in Business Intelligence: A Report From the BI Summit

Building Business Intelligence

Information Management Magazine, December 2007

William McKnight

Log cabins, ospreys fetching food out of the Rogue River and nesting atop cedar, fir and maple trees, Shakespeare Festival time and … business intelligence (BI)?! I’m obviously speaking of a different season from the winter in which you read this. This was summertime in Oregon, with reporting delays just due to procrastination on my part in writing about Scott Humphrey’s Business Intelligence Summit 2007.

However, it’s just as well because I’ve had a chance to ruminate on the many learnings I had and some of the things that I originated for the attendees. Unlike the dam, adjacent to the Weasku Inn where the event was held, which is going to be removed soon due to its lack of electricity generation, the discussions conducted there will continue to hold water for us attendees in our daily existences as consultants, vendors and press. I was privileged to be alongside Claudia Imhoff, Colin White and Jill Dyché as the featured presenters.

This was my fifth BI Summit, and it’s interesting how every year it serves as a leading indicator for what we all need to do our jobs well for the next year. As the summit topics shift, so does BI. Or maybe it’s vice versa. But regardless, consider how much master data management has crept into BI landscapes. Or the fact that we now look to apply BI opportunities operationally. How about the need to handle streaming, high-volume data, such as RFID data? Data warehouse appliances? In-depth reviews. Software as a service (SaaS)? Coming to BI - what’s the impact? And finally, data is so important that we need to govern it properly with high business involvement. These are burgeoning horizon opportunities as well as current needs that I address daily with my clients.

Jill asserted that there remains a great divide between IT and the business, especially in the area of BI. Two sources of the rift seem to be that IT is not teaching the business about IT and the proliferation of business-unit IT groups. These groups are allowed to communicate with the business more easily. However, these relationships can become intermediate gatekeeper roles, and other organizations have to travel through it to talk to that business unit. These setups have marginalized corporate IT and have become ways to attempt fixing problems organizationally, when in reality there are process problems.

Jill encourages engaging the business at the granular level of a discrete release of an application, making the business-IT conversation more concise, specific and refined, utilizing quick-hit, agile requirements-gathering sessions relevant to immediate results versus the broad-reaching requirements gathering of the past. Subsequent to that, continuing to slice projects into finer discussion points can develop more regular, consistent conversations with the business - necessary for success.

As a best practice, one client has IT people shadow businesspeople for two months in order to fully understand the business area they support. Sharing and cross-pollination of roles seems to work as a best practice process.

SaaS was an interesting and timely topic because several of the vendors present had SaaS offerings. SaaS has evolved significantly from the original application service provider model. Claudia pointed out that business process outsourcing (BPO) and managed service provider (MSP) are versions of SaaS.

Claudia mentioned that the advantages of SaaS are many, and much more is expected in this delivery model for BI in the future: 

• Vendors can support one platform and one version of the application,
• SaaS gives the vendor great visibility into how their customers are actually using the software,
• SaaS gives the vendor a predictable cash flow, and
• Vendors don’t get involved in “feature bloat.”
• Regardless, the most prevailing underlying problem with SaaS is the same as that with company-housed solutions - data quality – and customers still need to integrate SaaS application data with other enterprise data for other BI analytics.

Data warehouse appliances (DWA) are on the scene. Colin counted 10 in his research. I was informed of two more this week! So, that makes 12, possibly more in the market by the time you read this. Aside from the DWA analysis I and others have put out here at DM Review, some nonobvious areas of the DWA phenomenon were explored.

Page 1 of 2.