Author: Yogi Schulz
Ever notice how systems projects appear to be getting bigger and bigger? Once upon a time, any project with a budget over $1 million was regarded as a major project. Now a project needs a budget in excess of $100 million to attract much notice.
The media has reported on many monster systems projects including new air traffic control systems for the U. S. and Canada, the toll collection system for Highway 407 in Ontario and the stock trading system for the Toronto Stock Exchange.
What are some of the trends that have led to these monster systems projects?
Globalization is an increasing reality for many businesses. Unlike the local firms of old, the newly minted global organizations demand systems that span countries, continents and oceans.
As manufacturing continually seeks lower wage countries while engineering and design locate in high skill communities and marketing gravitates to the warmest climates, the demand grows for monster systems to keep track of business activity.
Once we bought our electricity, natural gas and telephone service from sleepy giants.
Now these giants are trying to learn ballet. What is their favorite weapon? Fancier systems that store more information about each and every customer in order to keep them from being picked off by the brash competitors pounding at the gate.
Companies like Bell Canada, Ontario Hydro and Enron are investing millions in monster systems to improve customer service, minimize billing screw-ups and generally make customers feel better about their service providers.
Searching for the Holy Integration Grail
Once upon a time, each department in a company operated a set of systems to manage its assigned function. Finance operatedfinancial systems; distribution owned shipping and invoicing systems; engineering coveted its CAD/CAM workstations.
Then one day, someone noticed that the same data, somewhat transformed, traveled throughout the organization. It was often re-keyed from computer printouts, produced in one department, into the computer systems of the receiving department. Various reports that purported to show the same data, swirled the numbers differently, didn’t add up and didn’t reconcile in accounting parlance.
Lo and behold, the neighborhood SAP sales representative had an answer. “Install our system,” he said, “these problems will disappear with our integrated solution.”
Integration holds out the holy grail of huge benefits in terms of lower capital requirements, shortened cycle times and reduced operating costs. Funny thing though, the integration monster project, designed to capture these benefits, is often the largestsystems project ever undertaken.
Voracious Appetite for Data
Western civilization believes that analyzing more data results in better decisions. This is baloney but nonetheless creates demand for monster systems projects.
Consider the billions that companies have invested in financial reporting systems during the past thirty years. Did these monster systems contribute materially to achieving the mission of the organizations or did they simply produce distracting detail for everyone to poke at?
For example, the grocery industry spends millions analyzing check out scanner data. What has all this analysis produced in the way of better products, reduced inventory carrying costs or heaven forbid, lower consumer prices? I bet the analysts can now accurately describe the weather conditions outside the store based on the scanner data. Wouldn’t it be simpler to phone Atmospherics Canada?
In the petroleum industry, engineers are convinced they must add last month’s production volumes to a 360 term time series of monthly production volumes before a work over decision can be made. Give me a break!
How many companies have monster data warehouse projects underway? Did anyone notice the recent article that suggested that Eaton’s data warehouse project contributed to its demise?
These trends are here to stay. I expect we’ll see many more monster projects. Before you embark on yet another monster system project, stop and ask yourself: “What were the lessons we learned from the last one?”
Often the lessons include:
- It’s easy to under-estimate costs and over-estimate benefits.
- Hardware and software vendors are prone to over-stating their capabilities.
- Our organization can not absorb change beyond a certain rate without serious cracks developing.
- When speed of implementation is achieved by reducing quality, it’s a poor trade-off.
- If we split the roll-out into multiple releases, the undesirable side-effects of change can be managed better.
Projects of similar size, which are being executed with less fanfare, include the overhaul of the systems portfolio at CP Rail, the implementation of SAP-based systems in major corporations and the introduction of natural gas and electricity trading systems.
As consumers we expect more.
Airlines will now accept orders for kosher or diet meals in addition to aisle or window seats. We expect airlines to not only keep track of our reservation but also our frequent flyer points.
Pizza take-out restaurants track our last few orders based on our phone number to speed ordering and pizza arrival.
We want catalogue companies to process our order for shipment the next business day. We expect that color, size and price will be unerringly accurate.
All these expectations require monster systems.
Moving into Operations
Most organizations have completed the application portfolio required to support the administrative aspects of their business.
Now, when organizations assess their systems priorities, they often discover that the most pain and the biggest opportunities exist in operations.
For example, dispatch systems for truck lines offer more benefits than enhanced maintenance tracking systems.
For petroleum companies, field data gathering systems can significantly improve production volumes.
For insurance companies, laptops for sales representatives with pricing and policy generation capabilities offer high returns on investment.
Achieving these benefits, often involves a monster systems project.
Where once postal mail was seen as speedy and reliable, fiber optic links now barely meet expectations for capacity and response times.
Systems projects are growing in size and complexity as we seek to exploit the capability and capacity of available bandwidth.
For embassy to embassy communications, highly secure, encrypted networks have replaced the diplomatic pouch with attached handcuffs and courier.
For military communications, satellite telephones and video links have replaced the signal flags made famous by Lord Nelson.
For the distribution of catalogues and pornography, the Internet has significantly reduced the volume of printed media.
Systems Development Tools
Today’s systems development tools can manage a level of application complexity in terms of the data model and the business rules that was unimaginable only a few years ago.
Have you ever compared an IMS schema to a relational schema? Notice how the relational schema contains many more tables, indices and foreign keys.
Development tools offer higher programmer productivity and better source and object control than ever before.
Why hasn’t the age-old application backlog disappeared in the face of all these improvements? End-users have ordered up more monster systems than even the newer generation of tools can deliver.
Capturing Data by the Gigabyte
New technologies for capturing data now meet the demands of even the most voracious analyst. As a result, systems which just index and manage all the media have become monster systems projects in their own right.
For example, satellite telemetry images now literally fill warehouse racks with data each month. Mounting tape cartridges to has become a full-time job for teams of sneaker-clad staff.
Quantum increases in seismic and log data recording in the petroleum industry have reversed the perception that the warehouse racks of yesterday would be obsoleted by the newer higher density media.
Medical images produced by CAT scans, ultra sound recorders and digital x-ray machines are creating mountains of bulky floppies in patient records.
Everything from workstations to super computers feels so wonderfully cheap compared to hardware prices of only a few years ago. This alluring observation leads to monster systems projects.
For example, we can roll out video conferencing to many desktops. We can distribute the dental plan application to every desktop. We can install high-availability servers for our key systems, whether they’re needed or not.
What we conveniently forget is that hardware is a small part of most systems projects total cost.