Business Intelligence Blog

Is a BI Project Different From Any Other IT Project?

onlyamodel — Wed, 27 May 2009 19:25:29 +0000

I’ve been reading a great deal lately about how BI projects are doomed to failure with one major study saying only 13% of all BI projects succeed. Obviously, this prompted me to inquire as to how can someone get a BI project in the 13% instead of the remaining 87%. Most of the pitfalls that I’ve seen are things that have plaqued software development for years.

First, most BI implementations, while they don’t always have to be, are large projects. Management is throwing a lot of capital at these projects, so they’re high profile and they usually touch all areas of a business. Are they as big as something like implementing SAP? Obviously they’re not, but they are in the same spirit in that they affect most of the departments of a company and there will most likely be some change and cultural management involved.

It’s important to understand the key strategic goals of the organization and to keep those in mind while crafting a BI solution, but at the same time you cannot forget about the goals of the BI users. Their goals may be as simple as let me get my job done quickly, so I can spend time with my family or as complex as how they can conduct their business while climbing mountains.

With long-running projects, such as a BI implementation, users might forget what they actually asked the team to create for them in the time it takes to craft a solution. In addition, business rules might have changed in that same timeframe. One way to resolve this issue is to go with the agile approach and demo the progress periodically while getting feedback from the user community. The same drawback to agile applies, which is setting project timelines becomes much more difficult. However, the main issue should be to deliver a quality solution and if a tradeoff between schedule and quality absolutely must happen, it’s always better to err on the side of quality.

Questions…

onlyamodel — Wed, 27 May 2009 13:54:49 +0000

What metrics are import (rank the metrics in order of importance)?
Who will be using the BI tools (are there cultural issues or fear of technology to deal with)?
How are new requests handled (need process for requests)?
What specific requirements are there (need requirements to deliver quality product)?
What is best infrastructure (software and hardware) for delivering current requirements and future requests?
How toI determine ROI of BI?

The questions here are designed to facilitate the implementation of the BI. Number 1 is designed to help keep the strategic goals in mind. Question 2 helps us keep in mind who we are designing for. 3 will put a process in place to make sure all requests are handled fairly and accurately. Question 4 gives us the requirements we need to determine if we deliver a quality product. Finally, infrastructure is import for the current and future success of the BI program. Number 6 is useful when reporting to upper management the benefit of their investment in BI.

Useful Links…

onlyamodel — Tue, 26 May 2009 17:44:32 +0000

http://biforbusinesspeople.blogspot.com/ - Talks about innovative solutions
http://www.ncc.co.uk/aboutncc/press_rel/Business_Intelligence_Implementations_fail_to_meet_objectives.cfm - Why 87% of BI implementations fail to meet expectations
http://it.toolbox.com/blogs/simplified-bi/why-do-bi-implementations-fail-8519 - Pitfalls of BI
http://peterthomas.wordpress.com/2009/03/11/gartner-sees-a-big-discrepancy-between-bi-expectations-and-realities-intelligent-enterprise/
http://www.itbusinessedge.com/cm/blogs/all/bi-sometimes-opens-can-of-itbusiness-worms/?cs=11450

Strategic Goals

onlyamodel — Tue, 26 May 2009 15:19:22 +0000

One thing that’s important to remember in BI is to keep the big picture in mind. The strategic goals of the organization should take precedence over things of lesser importance. Is the goal to increase sales with this information or maybe discover an untapped market? Perhaps better understanding of an existing market would be beneficial.

Excel Pivot Tables

onlyamodel — Wed, 20 May 2009 15:51:42 +0000

I keep seeing Excel being touted as one of the main end user tools for BI. As a developer I’m not new to Excel, but I’m also far from an expert on the subject, so I thought I’d see what the tool can do. I’ve used Excel Automation in .Net apps, created several queries for spreadsheets that import data and used Excel and the csv format for output on a number of projects. However, I’ve never used pivot tables, so after spending a couple of hours working with them, my conclusion is that it’s the equivalent of the SQL group by. I suppose users can use what they’re comfortable with and Pivot Tables do give a user familiar with Excel the ability to group by a column and perform aggregate functions like sum and count.

Microsoft Business Intelligence

onlyamodel — Tue, 19 May 2009 21:49:18 +0000

I’m anticipating a new project in which Microsoft Business Intelligence will be used. I’ve spent a good bit of time researching exactly what all that entails, and have discovered that it’s a huge umbrella of technologies, strategies and ideas. Any good sites, blogs or comments would be welcome.