Editor’s note: Joe Luedtke is starting the new year with a new focus for his column. Formerly he wrote Toward Pervasive Computing. This year, he will address ways small and medium-size businesses can effectively implement a business intelligence solution.

It’s really just a matter of simple math. The purchase and implementation of business intelligence software and solutions can be costly, very costly. The licensing alone for many business intelligence portals, ETL tools and enterprise performance management solutions are easily in the six figures. Even just a leading query and reporting tool can be well into the five figures for a set of licenses for a group of people. Building an enterprise data warehouse? Once you add together the requisite hardware, DBMS, modeling tools, querying and reporting tools, ETL tool, Web interface, consulting and support staff costs, you may now even stretch into seven figures.

While much has been talked and written about regarding the return on investment of business intelligence initiatives, little has been discussed concerning the barrier to entry that many small and medium-size businesses (SMB) face when staring straight in the eyes of a business problem that requires a BI solution.

Available Capital

If you’re not sure there really is a price barrier to entry for BI, let’s do a little math. There really is no clear definition of the small and medium-size business sector. While occasionally expressed in the number of employees (100 – 999), the SMB sector is typically expressed in terms of annual revenue. In this column, I’ll define the SMB sector to be comprised of companies under $500 million in size. With Gartner forecasting an average IT spend for an SMB of 3.2 percent of annual revenue for 2004, the constraints of the small and medium- size business start to come into focus.

With an IT budget of 3.2 percent of annual revenue, a $100 million company only has slightly more than $3 million to spend on IT, and that’s all of IT not just BI. Assuming at most a business intelligence initiative would get 10 percent (a very generous estimate) of the total IT budget for the year that company has only $320,000 to spend on their business intelligence solution. Now allocate two FTEs for database administration, report development, ETL processing and general system support and you’re down to probably less than $100,000 to spend per year on hardware, software, licensing renewals and consultants.

Looking at these numbers its clear why over 40 percent of companies within the SMB sector say they don’t need any business intelligence solutions. For many companies, the debate as to whether or not they could benefit from a BI solution or not is largely moot. They simply can’t afford them.

Half Empty or Half Full?

Okay, so I am taking the cynical view that the BI glass is less than half empty. If we treat it as half full, the good news is 60 percent of the SMB sector do believe they need a BI solution and either already have implemented one or have one currently under development. Yes, the price barrier to entry can be breached, it can be done effectively and it can ultimately produce bottom-line results.

A BI initiative within an SMB is not just a BI initiative from a Fortune 500 company on a smaller scale. The reduced scale of an SMB actually alters a typical BI methodology in a few significant ways:

Value versus Quality. An SMB will not necessarily buy the cheapest software available, but it will search diligently for the best value available. Name recognition and the vendor position on Gartner’s Magic Quadrant chart doesn’t matter nearly as much as the current value the software is capable of providing. In short, SMBs have no objection to shopping at a Wal-Mart or Target as long as they continue to get good quality at a low price. Crystal Decision may not beat out Business Objects or Cognos in Gartner’s "Completeness of Vision" category, but it may beat them out in value.

Bare-Bones IT. An IT staff of 1,000 can easily specialize and absorb new and specific technologies. However, it is another issue entirely if your IT staff is in the 2-25 range. For a small IT department, standards are even more critical than for a large enterprise. It’s really just a question of supportability. A small IT department will not create and staff a BI competency center. They will not select a J2EE-based BI portal as their BI front end if they are an all- Microsoft shop. They will, though, add BI development and administration to the workload of their existing development and administration staff.

Architectural Purity versus Supportability. The small IT staff within an SMB also translates into a reduction in complexity of the selected BI architecture. I once served on a committee in a Fortune 1000 company where we performed an analysis and proof-of-concept bake-off of Kimball’s and Inmon’s competing BI architectures. In this project, we modeled a portion of their proposed enterprise warehouse in both architectures as we worked to select the best architecture for the enterprise in this three-month project. On the opposite side, I once consulted with an SMB who created a single denormalized tab-delimited file of customer data and was perfectly comfortable calling this file their company’s data warehouse. In short, adherence to theoretical architectural doctrine may not be practical given the size of the IT department. Again, it is the value of the solution that matters most.

Minimal Use of Consultants. Consultants are used on a BI project within an SMB, but they are used much more judiciously and typically for very specific tasks. An SMB will be much more interested in hiring a consultant for the initial installation and configuration of a particular piece of software or the data modeling of their data warehouse than they would be interested in offloading an entire BI project to a consulting firm. Leveraging external expertise, knowledge transfer and training will be the key drivers in deciding whether or not to use consultants.

SMB’s face many of the same BI- related challenges as any large enterprise. While they are just different in scale, the scale difference narrows some possibilities while expanding others. The key problem is the initial price barrier to entry; however, there are also unique challenges that are a direct result of the constraints on IT funds: minimal IT staff, spartan use of consultants, limited BI/data warehousing skills, lack of an enterprise data warehouse infrastructure and significant more risk on projects (shorter ROI turnaround time, greater visibility and higher expectations).

Just how does a small and medium-size business effectively implement a business intelligence solution? Are the countless late night, brute force hours with MS Excel and flat file extracts the only answer? These are the questions we’ll be continuing to investigate in the coming months.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access