We regret to inform you that we will no longer be publishing Information Management. It has been an honor to provide you with the insights and connections to move your career forward. We wish you continued success on your professional journey and welcome you to explore our other titles at www.arizent.com/brands.

How to Get More Business Value Out of Business Intelligence

Luckily for most, business intelligence is in high demand. It’s a perennial top priority for chief executives — everyone likes to see a workforce making well-informed decisions that increase profits, market share, customer retention and satisfaction, etc.

This is good for the BI ecosystem; it’s a growth opportunity for BI practitioners and technology suppliers.

On the other hand, the fact that BI is a top priority year after year indicates that there’s some dissatisfaction with the progress companies are making on the BI front. In order to maximize positive perceptions of your BI efforts, business value creation should be a top priority when allocating BI development resources to projects.  

Before making BI value creation a priority, let’s define BI “business value.” BI business value is the costs reduced and/or the profitable revenue generated as a result of actions taken based on insights provided by BI.

If you buy into this, then the most straightforward way to increase BI business value is to increase the number of people gaining actionable insight from BI.

Here are three ways to maximize this:

  1.  Ensure new BI insights will be actionable.
  2.  Make new solutions a higher priority than upgrades.
  3.  Buy more, build less.

Ensuring Actionable Insights

This is a requirements analysis (profiling) task - actually, it’s really an art. When analyzing requirements, always ask users questions such as:

  1. What business function does the BI application support?
  2. How frequently will it be used by whom and by how many?
  3. What business activities will it affect?
  4. How will the output of reports, dashboards and scorecards be used, and what information is needed to act on new insights? For example, if inventories are out of balance (either high or low), you need to know when the next open purchase order for that item is due to be delivered, in order to push the receipt of the purchase order back if inventories are high or pull it in if inventories are low.
  5. Can you monetize the expected value of this BI solution? Will it help save costs? Increase income? How will it affect financial statements?

Questions 4 and 5, in particular, often go unasked. BI delivery teams are often preoccupied with technical feasibility and neglect to ask, “Why build it?” Business users with compelling answers to these questions should have their projects prioritized over others.
These questions may seem basic, but even within the last 12 months, I’ve encountered people who cannot answer them with respect to a potential BI project. For example, a data warehousing technologist within a financial trading firm in New York was ready to recommend the purchase of a new data warehouse database and commit resources to its implementation because it was capable of answering questions 40 times faster than their current DBMS and also able to query data as it is being loaded into the data warehouse for real-time analyses. However, he could not answer any of the questions above, and his recommendation was, justifiably, dismissed.

Make New BI Solutions a Priority

There are a lot of aging BI systems out there, data warehouses in particular. While it’s tempting to want to re-platform a data warehouse so that it can store more data and answer queries faster, doing so instead of solving unmet BI needs is risky. It’s like serving some people second helpings before everyone’s been fed.

Ensuring actionable insight is always the first rule of thumb when prioritizing projects — some re-platforming projects might rate higher than new apps, but carefully make sure they do before passing up the opportunity to equip new users with their first helpings of BI.

Buy More, Build Less

After 20 years of practice, undertaking BI projects is still as costly and risky as ever. Twenty years ago, the same was true of operational applications. IT was stuck in a rut of lengthy and expensive development projects creating custom ERP, CRM and other enterprise applications. What ultimately broke IT out of this rut was the emergence of commercial off-the-shelf (COTS) applications from then-young vendors like SAP, PeopleSoft, JD Edwards and others.

Prebuilt (yet customizable) COTS applications became a popular success because, relative to custom-built software, they provided:

  • A faster way to meet business requirements,
  • Lower ownership costs,
  • Higher quality as a result of testing from both professional QA teams and customers, and
  • More functionality based on years of enhancement ideas from many customers.

Suddenly, building a custom ERP application no longer provided the competitive advantage it once did. In actual fact, it became disadvantageous to custom-build a commoditized application while competitors implemented COTS equivalents ahead of you and refocused in-house application development resources on the creation of new, strategic software systems.
BI organizations should recognize that a similar revolution is at hand for BI. This is because prebuilt BI applications, delivered on-demand over the Web as software-as-a-service, provide extra benefits in addition to the ones listed above:

  • Greener data centers because it’s hosted and accessed over the Web, and
  • Easier funding due to the fact that budget approval may be easier for more flexible licensing options.

Making Better Build Versus Buy Decisions

Achieving a better mix of built and bought BI applications is not easy; buying turnkey BI applications runs against conventional thinking. The following notions are popular within BI organizations, and predispose many toward a decision to build, rather than to buy, new BI applications. At the beginning of every BI project, it is critically important to challenge each of these assumptions. Doing so should lead organizations down a path to satisfied, better-informed users regardless of whether the result is to build or buy.

“Our requirements are too unique to be met by a commercial BI application.”

: Some of your requirements are too unique. In most companies, there is no shortage of BI requests. The most productive BI organizations are the ones that recognize the difference between strategic and commodity BI requirements and use a mix of built and bought BI applications to meet them. Use the following rules of thumb to help differentiate commodity versus custom BI needs:

  • BI data is in COTS applications like SAP, Oracle, etc. versus custom built/proprietary sources.
  • You’re in an established industry like manufacturing, as opposed to a newer one, such as online gaming.
  • The project supports a relatively undifferentiated part of your business such as HR, regulatory, F&A or supply chain, versus R&D.
  • Commercial BI application software alternatives exist for your BI project.

“Using a commercial BI application will reduce our competitive advantage”

: Industry standardization of ERP, SCM, CRM and other applications from a small number of vendors like SAP and Oracle has not limited competitiveness. This is unlikely to happen with BI either.
Also consider that it can take many revisions for a BI application to reach maturity and deliver maximum insight to users. SaaS BI applications have overcome that and may provide more initial functionality than a new custom BI application. Taking a build approach can be disadvantageous if competitors are implementing turnkey solutions.

Are Ad Hoc BI Tools the Answer?

Only partially. If the way to greater business value is through a greater number of people equipped with BI applications, then one might assume that ad hoc BI/analytic tools for end users is a remedy to the backlog of unmet BI requests. Self-service, ad hoc BI is something companies should strive for, but it’s gaited by the availability of clean, integrated data stored in a data warehouse or data mart. On top of that, ad hoc BI tends to serve people better when supplementing a well-tested, predefined BI application and being used to allow users to go further from this base. Just putting data "out there" without really scrutinizing its suitability for BI purposes will not guarantee broader business intelligence adoption and value.

Satisfying an organization’s BI needs is a big undertaking. It requires companies to allocate their BI application development resources carefully in ways that maximize business value creation. To achieve greater success, BI teams should take some pages out of the playbook of their colleagues in operational enterprise software applications and start rationalizing proposed projects more conscientiously.

It should also be acknowledged that the build versus buy revolution will affect BI just as it did operational applications 20 years ago. Finding more opportunities to deploy quicker-to-implement, off-the-shelf BI applications can free up valuable BI development resources to work on more custom BI projects and increase the overall output of the BI team and the business value BI users generate.

For reprint and licensing requests for this article, click here.