My online series of articles is focused on the need for businesses to “get serious” about their approach to developing enterprise business intelligence (BI) and data warehouse (DW) capability. When pursuing this capability it is important to adopt a holistic view, followed by disciplined investment and execution. To develop the future vision for this capability, one should consider seven interrelated areas:

  1. Strategy
  2. People
  3. Process
  4. Metrics
  5. Applications
  6. Data
  7. Architecture

This column explores the key considerations of the Architecture focus area.

 

“A doctor can bury his mistakes but an architect can only advise his clients to plant vines.” - Frank Lloyd Wright

 

I’m now writing my 11th article about the things that need to be considered in order to establish and execute a strong enterprise BI/DW capability within a company, but this is the first article focused on technology. This may come as a surprise that I’ve managed to avoid “tool talk” all this time. This is not meant to indicate that tools are not important to the success of a BI program, rather it indicates that they are only one piece of the overall puzzle. Unfortunately, some IT people and business people put so much emphasis on tools that they ignore all of the other things that can ensure success in their BI endeavors.

 

Take a Look at Your Footprint

 

In many companies, it is common to find BI solutions within IT and/or business groups that use a variety of competing tools. It’s also common to find that tools that are important to success are actually missing from the picture. As with most things that impede success in enterprise BI, these technology footprint issues are typically caused by siloed thinking and funding occurring out of context with any long term vision.

 

Having multiple competing tools that provide basically the same functionality presents several challenges:

 

  • User communities overlap. When the redundant tools are user facing - like reporting tools or OLAP tools, as each BI solution’s user community grows users will find that the user communities will begin to overlap. When this happens, the end users will start to become frustrated that they have to learn to use so many different tools. It will also result in less productive use of these tools by the end users due to the learning curves.
  • Overall company software costs will be higher due to the purchases of the redundant software plus the ongoing software maintenance costs. Additionally, the software is typically more expensive on a per user basis when purchasing in smaller volumes.
  • Overall company hardware costs are higher due to the need to run the software on different servers
  • Overall labor costs for support are typically higher because you end up having more people supporting the redundant software tools and BI applications than you would if you had a narrower footprint.
  • Overall labor costs for new development are typically higher because you end up needing to have more people in order to provide development coverage for the multiple tools. It also makes for less effective utilization of these people, since there may not always be enough work to keep each of the many “tool experts” busy. Or worse you may end up finding new projects for people to do - just to justify someone’s existence - even if that new project really isn’t important in the grand scheme of things.
  • Managing security is difficult. A more subtle, but also important consequence of having multiple competing tools is that it becomes very difficult to manage security in any common way. For example, it would be very difficult for a “data owner” to determine who has access to their data because it may be distributed through multiple reporting or OLAP tools that each have their own security authorization method, thereby reducing people’s trust in the collection of BI applications.

Having gaps in your architecture and your “box of tools” also presents challenges. When lacking the right tool for the job, people will tend to compensate by using tools that were never meant for that purpose. This can cause problems such as:

 

·      When a BI application needs to be scaled to a larger number of users or larger data volumes than the retrofitted tools can support

·      When business needs change and the tools used to create the original solution are not flexible enough to support the change and end up having to go back to the drawing board.

·      Usually when people are “making do” with a tool that isn’t fit for the job they end up spending more money in labor time than the tool would have cost in the first place to get the same job done.

 

Pick Some Standards and Stick with Them

 

If you are having any of these technology footprint challenges, the recommendation would be to first try to standardize on one tool for each of the primary categories of BI software that you need.

 

The most basic types of tools needed to be successful in BI/DW include things such as:

 

  • Relational DBMS,
  • Data modeling tool,
  • ETL tool, and
  • Relational Reporting tool.

The next step up in capabilities would include things like:

 

  • Multidimensional OLAP DBMS,
  • OLAP User Interface tool, and
  • Dashboard/Scorecard tool.

More advanced capabilities would be provided by these types of tools:

 

  • Data Profiling tool,
  • Data Cleansing tool, and
  • Master Data Management tool.

Additionally, it’s a good idea to avoid the tendency to join the “tool of the month club.” IT and business teams can slip into a pattern of constantly evaluating new tools when the current tools are - for the most part - getting the job done. In my experience all tools have some tradeoffs that you end up having to get used to, but the solution doesn’t usually need to be getting a new tool, because the new tools are going to have their own issues too. Switching tools should be done with a careful understanding of the impact on existing applications, the costs of software, hardware, and training, and the potential inefficiencies of users and developers.

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