A Better Warehouse For Better Business

Vendor Efforts Are Paying Off In Expanding Data Warehouse Capabilities.

The way some data warehouses behave, it's no wonder we keep re-inventing them. Withadvancements from vendors adding up, however, reliable end-to-end systems may now bewithin reach. Take a look at developments that are unlocking true, enterprise-levelcapabilities from data warehousing solutions.HP Professional Staff

ata warehouses wire us up for something like e-intelligence," says Rick Millem, HPData Warehouse Marketing. "As we currently see it, e-intelligence is an evolution ofbusiness intelligence and leverages the Internet while drawing upon the datawarehouses/marts and the advanced search algorithms of new data mining tools." Italso relies upon universal access to the Internet, intranets and extranets through whatMillem refers to as "enterprise information and discovery portals."

An example of this is TopTier Software's (San Jose, Calif.) Enterprise InformationPortal. With user interfaces made to display all information as though it's from the samedata source -- combined with back-end functionality that allows the data to be comparedand analyzed --vast amounts of previously underutilized data can be refined intoinformation for newer, more personalized sales, marketing and distribution activities.

A Thinking Man's Fortune Teller

To gauge how important another aspect, data mining, is becoming consider Oracle'sacquisition in June of Thinking Machines Corporation (Bedford, Mass.). As PC Week Onlinereported, that was Oracle's way of fast filling a gap in its data warehousing strategywith Darwin, Thinking Machine's server-based data mining product.

Darwin, like other data mining products in general, uses algorithms for exploratorydata discovery. Sales and marketing executives have previously sought these third-partyapplications in order to identify product sales affinity and cross-selling opportunities,for example; or to monitor repeat customer sales or fraud activities. These applications,which are luring a critical mass of interest among business executives, are becoming abreed of functionality that data warehouse solution providers are providing integration toor acquiring individually for their warehouse suites.

To date, data warehouses have provided ways for analysts and researchers who've beentrained in these tools to see and report on historical trends. With the advent of OnlineAnalytical Processing (OLAP) technology, managers as more casual users have at last,efficient ways to model a wider variety of "What-if?" scenarios and play outmarketing strategies to plot their next moves.

What if inventory levels drop, but aren't noticed and reported for a week or more? Whatif critical shipments don't arrive as expected? These kinds of questions have prompted yetanother new class of tools -- business alert systems -- for addition to the data warehousestrategy.

Staying Alert To New Trends

Business alert systems monitor pre-set thresholds on a wide variety of data sources andare ever ready to send a signal to the appropriate person once a threshold is crossed.Rather than waiting for reports to point out trends and opportunities, alerts keepbusiness managers abreast of the vital operations and allow them to act more responsivelythan was previously possible.

The original dilemma of data warehouse developments -- that they're never finished --isn't something that a pessimist should describe as a warehouse half-empty. The opensystems environment has made optimists of many and proven its worth by affording room tobuild further and providing streamlined, highly-enhanced business operations.

Data mining digs deeper and faster for information that's been developed and tuckedaway. Improvements in this area will prevent companies from failing to turn transactiondata into business intelligence. New data mining applications put powerful, new algorithmsunderneath user-friendly interfaces. In conjunction with familiar tools like Hyperion's(Sunnyvale, Calif.) Essbasse or PowerPlay from Cognos (Ottawa, Ontario), the data miningtools dig deeper without overly relying on the analytical skills of users.

Quick return initiatives will be rewarded quickly with tactical time/opportunity-drivenbusiness intelligence capabilities for ERP users for multiple-source data warehouses.Joint work being done by Platinum Technology (Oakbrook Terrace, Ill.) and Ernst and YoungConsulting that claims to provide customers with rapid deployment of strategic datawarehousing for ERP decision support is a good example of how data warehouse managers areable to introduce high-return systems as needed.

Calls To Action

Highly active industries such as telecommunications will get industry-specific extracttechnology, data models, methodology and architecture. This solution turns call detailrecords into business intelligence for campaign management, churn and fraud reduction,marketing and sales.

With some warehouses bursting at the seams, multi-terabyte warehousing is refined. HPand Oracle recently announced a terabyte-plus initiative for customers requiring datawarehouses in the mega-terabyte range.

With vendors working in partnership with customers and consultants, IT managers willalmost certainly find the waters of data warehouse design and development significantlyless murky than they used to be.

There's still every reason to move carefully. Nevertheless, the good news is that datawarehouse programs can now build from proven best practices and tightly integratedtechnologies.

The resulting solution, and you can call it an enterprise-capable solution, willincorporate significantly better generations of front-end tools that solution vendors havedeveloped for a level of synergy that was the potential and promise of earlier systems.