Blog: Craig Schiff

« February 2006 | Main | April 2006 »

March 22, 2006

How NOT to Select a Performance Solution

Ever been to an IT chat room or forum? They are great places to get solid answers to pressing technical challenges. Leveraging the combined experience of all participants in the room or on the distribution list is probably one of the best ways to get a quick, authoritative answer to a detailed syntax or command question that the vendor failed to adequately document. However, I have recently seen these forums used for a purpose that I don't believe they are well-suited for: BI and/or business performance management vendor selection. The typical query runs something like this: 'we are a large, complex bank and we are looking at Microstrategy or Cognos to analyze our data, which would you recommend?' Just in case the problems with this approach aren't already obvious to you, let me spell them out.

First of all, most people in these groups use screen names so you don't know who you are really talking to. When getting technical help that doesn't really matter. When making a sizeable purchasing decision it matters a lot. Is 'BIGuru' a salesman for Cognos who will promote his product or a disgruntled ex-employee who will have nothing nice to say? The next issue is that to properly select a solution that you are going to live with for some time to come you really need to delve into your specific detailed requirements and match them up against the product choices and their capabilities. While product A may have worked quite well for someone else at a bank, it may not work well at all for you. Your requirements, at the detailed level, are most likely going to be different in some important ways from the people responding to your question. Lastly, let's take a look at what we are attempting to do here. We are looking to buy a BI or performance solution that will be strategically important to the company, probably cost hundreds of thousands of dollars when all is said and done (software+services+staff time), and live on for 5-7 years. Is this the way to make a decision of that magnitude? Of course not. You need to go through a comprehensive, structured process to determine your detailed requirements and evaluate all the viable vendors for this solution. An IT chat room can be one small step in a thorough due diligence process, but it certainly should not be the only step. I wonder how the senior executives in these companies would feel if they found out their BI/BPM experts used this as a primary means of selecting vendors?

Share:

Interest in Master Data Management Runs High

All of us so-called 'industry experts' have been talking about this topic for a while. We may call it master data or meta data or enterprise dimension management, but it all means the same thing. The question in my mind has been 'does the end user community see the same pressing need that we do?' Two recent data points have answered that question for me. I was asked to speak at a webcast on this topic, specifically focused on the chart of accounts dimension. Well, as it turned out the event was over-subscribed and we needed to schedule a second live run of the same webcast to handle the overflow crowd. In addition, we just completed our annual BPM Pulse Survey which asked nearly 500 respondents to rate the importance of a list of up and coming features of business performance management (predictive analytics, vertical solutions, etc.) and meta data management received the highest ratings. MDM/EDMs time has arrived.

Share:

March 2, 2006

New Decision Support System: Crowds

At a recent business performance management conference they had James Surowiecki as a speaker. He is the author of a book entitled 'The Wisdom of Crowds'. The fundamental pricnicple is this: if you take a diverse group of individuals and aggregate and average their input, you will get a better answer than just relying on any one individual, even an expert on the topic being discussed. I of course was skeptical at first, but he provided many examples to prove his point. The theory appears to work on everything from determining how many jelly beans are in a jar to forecasting future outcomes of today's business decisions. His hour-long speech from that event is now available here as an MP3 file (it's large, best to right-click and save target as). You can decide for yourself if it makes sense. If he's right, we probably need to change the way most business decision making is done, regardless of the underlying software and data warehouses that provide the information needed to make those decisions.

Share: