A few years ago, the chairman of a large manufacturer found out that his organisation was spending $1.25 billion on IT on a yearly basis. This amount was high enough for him to hire an outside VP who was to investigate how this happened, and report to back to him, directly.
After several weeks of investigation, back came the answer on how it happened:
"One excellent decision at a time"
It turned out that the only message IT ever received from management was: "Don't stand in the way of growth". Since growth at this company had been astronomical nobody even had time to think about a structured IT policy, or keeping cost under control. Since revenue outpaced IT spend by an order of magnitude, nobody outside IT ever cared.
Only until the growth rate of the company leveled off and cost-cutting became crucial, did management pay full attention to the facts.
This is by no means the only organisation where IT cost have spiralled out of control. Large organisations invariably spend over 90% of their IT budgets on maintaining their current state. (For the above manufacturer, it was 96%). Less than 10% is spent on creating and delivering new services.
How is this possible, on such a large scale? By "not getting in the way of growth", until the structure collapses on itself. Sometimes literally, as evidenced by a number of large scale IT outages over the summer.
Think, if you will, about the way medieval towns have been constructed. Completely organic to the point of becoming unmanageable. In order for them to stay alive and prosper, structural, fundamental changes have been made over the years, such as adding underground railway networks, zoning laws and building codes. This has allowed them to grow while keeping their inhabitants happy and productive. The medieval towns who have not changed their infrastructure are now either museums or ruins.
But this is a hard problem to solve. Under the various pressures of a competitive environment, no one likes to point out that in order to reach the next growth level, first a fundamental change is needed to the IT. Particularly because large scale IT projects do not have a stellar track record of delivering. So organisations keep tinkering with the existing systems, adding more complexity along the way. Which causes the cost to rise even further, but also increases operational risk tremendously. The more moving parts your engine has, the more difficult it is to keep them all running in sync.
Not until the problem reaches CEO level, or otherwise catastrophic proportions, is fundamental change possible. And for some organisations, by that time, it may be too late.
The situation can be prevented. A first step is to get a board level review of the current state of the IT.
This review will serve as the eyes and ears of the board into the state of IT. It will demonstrate the cost needed to sustain the operation, and whether internal or external systems operator can guarantee a certain service level. It will show where the systemic risks are. And it will allow the boards to see whether they have a problem they need to address, before it might be too late.
What would an IT review look like? First we need to scope the review to look at strategic transactions for this organisation. The review could for example look at processing a shipment (if you are a logistics organisation) or billing (if you are a utility company or a telecommunications company) or financial transactions (if you are a bank) or any type of transaction that may appear risky.
Then answer the following questions for the last three months
- transaction volume
- a list of distinct systems needed to process these transactions
- for each of these systems, number of correctly processed transactions, versus number of non-correctly-processed transactions
- for each system, the legal entity that operates the system, and the number of transactions it is guaranteed to process correctly. Barring that, any other guarantee this legal entity gives in operating the system.
- for each system, the cost of operating the system
- for each system, the availability over the last three months
- for each system, the number of problems (issues) reported over the last three months.
This allows the assessor to create a high level, but accurate picture of the chain of information systems needed to process transactions. And whether these systems haven acceptable operational quality level (through number of problems reported), whether these systems have an acceptable minimum quality level (through operator guarantees), and whether these systems have a sustainable cost level.
About the author: Tobias Kuipers is CTO of the Software Improvement Group