Welcome to Office Zealot Sign in | Join | Help
Where is the sense of financial discipline?

A client of mine recently asked me what I thought about the hardware requirements of an application the client is considering building. In a nutshell, the client wants to build a departmental management reporting application that will likely support about a dozen to two dozen occasional users. The hardware/software estimate to support this as determined by the client’s IT department: $103,000! No special software here – just Windows Server and SQL Server 2000. Basically they designed a highly available two-node cluster for failover purposes. With SQL Server 2000, you need to get the Enterprise edition to support clustering. One way to save a lot of money here is to wait a week and get SQL Server 2005 Standard edition (licensing permits a two-node cluster for failover purposes).

I’m mainly distraught by the over-engineered hardware requirements. In my opinion they designed about a four 9’s environment (~99.99% uptime) for an application that requires only one 9 (~98.9% uptime).  The difference in annual downtime between 98.9% and 99.99% is about 95 hours (out of 8,760 hours in a year). They should be able to put together an environment for this application for about $25,000.  Is 95 hours really worth $78,000 for a departmental management reporting application? If so, how do you justify their current situation which is next to nothing? That’s right – at the present time they have no way of rolling up sales estimates from their partners other than physically opening up and adding each of about 200 spreadsheets. If 95 hours is worth $78,000, then shareholders should be up in arms that they have allowed the current situation to exist for years.

These types of situations cause what many IT departments are hoping to avoid. Rather than wrapping up business processes and data in a more formal environment where IT can keep the pulse of the computing needs of the organization, it forces departments to develop applications on their own, under the radar of IT. These are the homegrown solutions that IT folks roll their eyes at and groan. But give managers some credit – faced with $100,000 just to get out of the gate (no development costs yet) or free (using Access on a fileserver), many managers who can live with occasional downtime are going to go with free. As a result, IT shops are faced with increased Access proliferation, not less.

Some of the more successful corporate IT shops I have worked with embrace a strategy of choice. They provide departments with options and let managers choose their level of availability. They let managers choose their platform. They provide “incubator” environments for prototyping or for smaller applications. They recognize that for every manager who doesn’t have the knowledge to choose for themselves, there is another one that does. They provide guidance where needed but otherwise they provide quiet enablement. These IT shops are viewed as trusted advisors rather than adversaries. They enable the creation of value rather than hinder it. Here’s to the creators of value!

Posted: Friday, November 04, 2005 8:23 AM by hansen

Comments

Anonymous said:

This problem sounds so familiar.

I have a client with 3000 retail locations. Each retail location sent in an excel workbook each month (36000 Excel workbooks a year). The store data was then cut and pasted into a giant Excel workbook without any automation.

I suggested a 100 hour project to standardize the reporting and automatically deposit the monthly info into a SQL database. The system would easily facilitate a very flexible reporting structure using Excel and ODBC including the generation of “same store sales” reports. I actually built a prototype in 3 days to demonstrate feasibility. This was June 2005.

Ok, so the client goes out and purchases Essbase and embarks on a massive project. This is mission critical stuff and “can’t be trusted to Visual Basic, SQL and Excel”. So far the cost has been about $200,000 and it still does not work. EssBase has some “hoky” program language and the Crystle Reports reporting tool. Have you ever tried to report financial statistics with a “list maintenance program”?

So, I get a call on December 10 - “Gotta have something that works before the January 31 year end. Can I get them over the ‘hump’ until they figure out how to make EssBase work!” FAT Chance, I’m a CPA and have tax season to contend with.

Go figure?
# December 15, 2005 3:13 PM
Anonymous comments are disabled