By the end of this year, worldwide investments
in business intelligence (BI) will exceed $16
billion. The promises of these investments
are manifold. In theory, more timely
and effective decision-making should translate
into more competitive strategies, greater market
share, improved customer satisfaction, more
productive employees, better control and oversight,
and more accurate planning, budgets, and forecasts.
The real-world results often fall short.
There are a number of misconceptions and factors that stifle the payoff of a BI
initiative. Among the worst offenders, or worst practices, are the following:
"BI is just a data warehouse."
A data warehouse offers structured business information. In and of itself, it does not offer business intelligence.
It is only one of many critical building blocks.
Equating a data warehouse with business intelligence
belies the very reason you implement BI solutions
in the first place, which is not only to structure,
store, and access information, but to interpret
and actually act upon it for the types of benefits
I just cited. True BI scenarios incorporate
much more than a data warehouse and a slick
UI. They work in lockstep with business process
management, collaboration tools, and business
applications. Tying analytics back into action — that's what BI is all about.
"Good reports = Good BI."
Many who design and implement BI strategies
are lulled into a false sense of security,
believing that good reports are indicative
of good BI. Building and delivering good reports
is only one facet of BI; it is by no means
a complete picture. Dashboards, alerts, balanced
scorecards, and KPIs are also highly effective
means for communicating business insights,
and are means that are becoming more popular
given the trend toward "exception-based
"BI ends when the report hits the executive's inbox."
If you want to gauge the efficacy of your
BI activities, take a look at what happens
once a report reaches your CEO or your LOB
executives. I see it time and time again. An
important issue warrants executive review and
the CEO asks for a report. The report is run
(or in some cases, created at great expense)
and sent via email to the CEO's inbox. The inbox offers no inherent support for collaboration among key stakeholders, no context upon which to draw meaningful conclusions, and no simulation or modeling capabilities — in
summary, no actionable intelligence. And most
glaring of all is the lack of tie-in to processes
and applications. BI should be a catalyst for
business transformation, an enabler of smart
business. That CEO should be empowered to make
better and smarter decisions (ultimately affecting
the bottom line of the organization), and to
quickly initiate process changes where recognized
to be advantageous.
"One size report fits all."
We all understand that interest in a specific
type of data — say, sales figures — doesn't
mean a one-size-fits-all report is appropriate.
A marketing manager may want to review sales
figures by promotional campaigns, a sales manager
by sales rep and product, and a purchasing
manager may want to focus on trends over time
in order to avoid out-of-stock situations.
And yet, many companies resign themselves to
having a data warehouse administrator run a
report and having the users decipher it on
their own. Or worse, the administrator (without
the relevant business understanding) runs numerous
reports tailored to what he or she perceives
are the needs of the audience. Inflexible,
impersonal BI such as this forces users to
make compromises and hinders the quality of
insights to be gained.
"Six different tools (and six different answers!)."
With the advent of Sarbanes-Oxley, acute concerns
about timeliness, accuracy, and consistency
reach well beyond the IT department. Internal
controls and trustworthiness of financial figures
are key concerns for every CEO and CFO who
has to attest to a financial statement. Compliance
can be difficult or easy to obtain, depending
on the route you take. Isolated data warehouse
implementations take you down a difficult and
expensive path to compliance. And it's
a path marred by risk. You extract information
from a data warehouse, manipulate it with external
tools, and then return the revised data to
the system. There is no single version of the
truth, and data modifications cannot be traced
"To get the best, buy best-of-breed applications and tools."
One of the myths that enabled best-of-breed
environments to take hold in the 1990s was
the notion that they offered flexibility. It
didn't take long for both the
business and IT sides of the house to discover the fallacy of this claim. It's hard (if not impossible) for a business group or executive to get to one version of the truth when different organizations take a different approach to BI. Moreover, manual integration of multiple products exponentially compounds IT overhead and expense. Modifications, software patches, updates, upgrades, and
integration exercises — basically, day-to-day dealings for every IT department — wreak havoc on a so-called "best-of-breed" BI landscape. Vendors don't
coordinate the release schedules with one another.
They all issue updates at will, leaving IT organizations
to administer change to one system, then revisit
every other system that is impacted by that change.
A single, integrated, and robust technology platform
from a proven vendor, not a best of breed, is
actually the key to flexibility.
"Analytical and transactional applications: different strokes for different folks."
Many BI initiatives operate in isolation. They offer reports, but are either not able or were not configured to track, measure, and refine operational execution. But the very premise of business intelligence is to assemble data for analysis that guides subsequent actions! It is absolutely essential that the BI solutions interact with operational systems, offer monitoring and feedback mechanisms, provide timely (or preferably real-time) alerts, and align business insights with corrective actions and process improvements.
The 7 Best BI Best Practices
The seven best practices to emerge in the BI arena are summarized in Figure 1. At the core of every one of these best practices is an acute requirement for a common, open, and unified technology platform. In this regard, SAP customers are ideally situated because SAP NetWeaver is precisely that.
|Run business processes and BI activities on one common, integrated technology platform.
||Remember that "Business Intelligence" starts with business, not technology, and that business is in constant flux. Run BI activities on the same platform that runs your business processes, and you can readily achieve and maintain alignment between them, even as business conditions and requirements change. In fact, much of this alignment (that which is not wholly unique to your business) should come prefabricated. This is certainly the case with SAP business applications and the SAP NetWeaver technology platform.
|Don't reinvent the (technology platform) wheel.
||The trend line in the BI market mirrors that which took hold in the ERP market: utilizing preintegrated technology suites or platforms has proven to be less costly, more flexible, and capable of supporting more robust BI activities.
|Foster interaction and cooperation among BI project participants early and often.
||BI projects draw upon a broad-reaching mix of skill sets. Communication and cooperation among business, portal, data warehouse, development, administrative, and other teams are streamlined when one platform is in place and all project participants speak a common language.
|Think big, start
|Starting with a focused effort, then iterating and refining a BI solution, optimizes its value. An integrated platform allows you to stay open and extensible (for example, you can activate features when the time is right).
|Invest in technology that is standards-based, open, and backed by a vendor that will be around for the long haul.
||Choose a platform that adheres to open standards, supports integration with third-party offerings, and has the size and success to run certification programs for a wide variety of products.
|Make data quality a top priority.
||Much rides on the integrity of the information and conclusions drawn from BI projects. An integrated platform significantly reduces the chance for data contamination and inconsistency.
|Architect for your current requirements and budget, but be acutely aware of the need for downstream growth and
||Strike the right balance between current budget constraints and inevitable downstream growth. Here, a common, scalable platform becomes a key asset.
| The 7 Best "BI Best Practices"
Harnessing the benefits of BI requires broad-reaching technical capabilities along a number of fronts. You obviously need extraction, transformation, and loading (ETL) capabilities, integration brokers, and data-modeling tools. You need a portal to enable access to information and analytic applications. Security is certainly a huge factor for every IT organization these days. And cutting across all facets of a BI endeavor is the need for integration with business processes and applications. These capabilities are all integral to the SAP NetWeaver platform.
When these technical assets are at your disposal, BI can take flight. You can use business process management in conjunction with analytics. You can wield the portal for broad-based delivery of content, be it reports, alerts, or KPIs. You can tap into collaboration tools so that people can discuss the contents of a report, reevaluate findings, and question conclusions. You can readily tie analytics back into transactions. And you can create processes around analytics, for example, via ad hoc workflow definitions and guided procedures.
The Breakthrough BI Best Practice
"Speed" and "BI" are not terms often heard in the same breath. This is a misconception. If you have the right
platform (and again, SAP customers do, with SAP NetWeaver), you can achieve high-performance BI even when contending with huge data volumes and large numbers of queries. In its next release, SAP NetWeaver leverages scalable and distributed attribute-search technology, vertical decomposition, horizontal partitioning, smart compression, and in-memory processing. You can expect speeds that are orders of magnitude, not just gradual improvements, over what you have now.
A New Era in BI and the Demise of
Nowhere is the term "Powered by SAP NetWeaver" more apt than BI. True business intelligence compels us to investigate, integrate, and innovate; to be quick and confident when we take action; and to ensure data integrity, consistency, and alignment with business objectives the whole way through. You can't do these things when wrestling with multiple platforms. You can with SAP NetWeaver.
Look out across the SAP customer and partner base (including the partners you see profiled in this Special Feature), and you see a new era in BI take root. Long-standing tradeoffs with regard to ease of integration, flexibility, performance, and TCO go away when you migrate BI activities to the SAP NetWeaver platform.
back to top