Some Short Stories
by Bill Inmon
Originally published July 30, 2009
The other day at a conference, I met a gentleman whose company specializes in building data warehouses for customers of JD Edwards. He was a very nice person and he had some interesting stories to tell. So we chatted before my presentation.
Now people come up to me and chat about their products and services all the time. And if I have the time, I enjoy hearing about what is going on. I learn a lot from listening to people. And I like to be polite. But this gentleman was unique. You see, his company was building data warehouses for customers of JD Edwards. And this holds a special place for me.
Years ago, I happened to be working in Denver, and our offices were just a few blocks from JD Edwards. We were neighbors. This was in the very early days of data warehousing and the tools for ETL and monitoring data warehouse activity were newly emerging. Corporations of all sizes were discovering that they needed data warehouses. I spent my time talking to people and I told them all of the things that data warehouses can do for them. Selling an entire industry is a long and protracted effort.
At that time, JD Edwards was a large and successful software company. The manager or president – I really don’t know who – was famous for saying, “If you have JD Edwards, there is no need for a data warehouse.” He went on to say, “There will be data warehouses at JD Edwards over my dead body.” For whatever reason, there was fierce animosity toward data warehousing by JD Edwards. In fact, there was so much animosity that at noon when people from JD Edwards came out to eat lunch, they would not share a table with me – even though our companies were very close neighbors. I was the “data warehouse guy” and I was treated as if I had some strange disease.
Well that was years ago. JD Edwards was bought by PeopleSoft. Then, PeopleSoft was bought by Oracle. Most people involved have gone on to other things by now. So here comes this company that is building and selling data warehouses to customers of JD Edwards, and appears to be doing quite well, thank you.
You may think that the moral to this story is that I got the last laugh because I was right about architecture and the top manager at JD Edwards was wrong. He just didn’t understand the first thing about architecture, and his company missed a major opportunity. Had JD Edwards understood architecture, they might be a strong independent software company today (as they once were). It is possible to interpret this story that way.
But there is another interpretation. That interpretation goes like this. Once there was a guy who ran a company and didn’t understand architecture at all. There was another guy who really did understand architecture. And today, the guy that didn’t understand architecture is retired, driving a Porsche and counting the sunsets in Puerto Vallarta; and the guy who did understand architecture is still working away at making a living. Maybe it isn’t good for one’s health to understand architecture.
Along the same vein, there’s another story about architecture. It comes from the same geographical locale as the JD Edwards story, although, in this case, the company name shall remain anonymous.
Recently, I was called in by managers of a well respected company to talk about their need to build a data warehouse. There was the recognition that there was no reliable data in the corporation. The organization was running on spreadsheets, and each department (and in some cases, each sub department) had its own pet spreadsheet. No two spreadsheets were remotely the same. Making decisions was something akin to throwing darts at a dartboard and seeing which part of the target the dart had randomly hit. The managers were concerned about the ability of the corporation to make fiscally responsible decisions in such an environment. They called me in to talk about data warehousing and how data warehousing could help create reliable, believable data.
Soon, we were making plans for the work that needed to be done for the data warehouse.
Then, things went silent. I didn’t hear from these managers again for a while. A month or so later, I got a phone call. It seems that a new top manager had been hired and had cancelled the data warehouse project. The manager had had previous experiences with data warehousing that were bad. In a previous job, a large consulting firm with impressive credentials was brought in. In fact, the large consulting firm had never before built a data warehouse successfully. The consulting firm decided that there was a lot of money to be made by this new fad of data warehousing. So, they decided to cash in. It was too expensive to go out and hire people who had really been building data warehouses. And besides, the large consulting firm already had proven methodologies that worked quite well for SAP implementations and the development of large operational systems. In addition, the large consulting firm had strong ties to the manager. They often went to fine French restaurants where the large consulting firm treated everyone to expensive Bordeaux. So the top manager selected the well known consulting firm to build the data warehouse.
After two years and $5,000,000 in consulting fees, the consulting firm admitted that it had to start over with the design because requirements for analytical processing kept changing. The methodology that the consulting firm was working from mandated that requirements be gathered and stabilized. Indeed, requirements were gathered, but as soon as the requirements were gathered, new ones appeared and ones already captured changed. There simply never was any stability of requirements. It was worse than herding cats. It was herding a whole roomful of cats, with a few mice thrown in.
The whole data warehouse effort was declared to be a fiasco.
So the top manager changed companies and vowed to never again build a data warehouse.
There is a certain irony here. The original pioneers of data warehousing told people that there was a certain way that a data warehouse had to be built. And thank goodness a fair number of people actually listened, because there were a large number of data warehouses that were successfully built. But there were a lot of other people who either didn’t listen or wouldn’t listen. And this is where the data warehouse failures came from. People were doing the very things the data warehouse gurus told them not to do, and when there was failure, data warehousing got the blame.
I told my friends (the managers at the company in Denver) to wait until the top manager moved on again and we would talk to the next top boss about continuing the data warehouse effort. Because by ignoring architecture and sticking his head in the sand, the new top manager has a predictably short life span.
Maybe the water or air at that particular location in Denver has something peculiar in it. JD Edwards bit the dust and the manager at this other company predictably is not going to last long. When does the day come when managers finally understand architecture and its implications?
SOURCE: Some Short Stories
Recent articles by Bill Inmon
Copyright 2004 — 2019. Powell Media, LLC. All rights reserved.
BeyeNETWORK™ is a trademark of Powell Media, LLC