18 April 2006

System governance: less is more

By Andrew Clifford

To gain the benefits of system governance, you only need simple materials, processes and tools. Anything more complicated would get in the way.

At Metrici, we have spent the last year defining our vision of system governance, and developing materials and tools to support it.

Sometimes I wonder if we should have done more. We could have provided more comprehensive materials, perhaps aligned to industry standards, such as ITIL, or COBIT, or the Zachmann Framework.

And we could have defined detailed processes for system governance, perhaps embodied in some sort of process management tool. We could have developed an advanced "workbench" tool, that brings together other related functions, so that the users never have to leave the tool.

But we have not done this.

We have developed some materials. They embody best practice built from a few decades of experience, but they are not very detailed. I would not claim that they are more than other experienced professionals could put together.

We have outlined the ways of using system governance, but we have not defined any overall processes for a system governance methodology.

We provide a simple tool that holds the information you need for system governance, and performs analyses. But it does not do anything else. We would expect our users to use other tools at the same time.

Of course we could have designed our products differently. But we kept them simple because of problems we have seen with other products.

Materials have to be understood. Complicated, in-depth materials, twisted to align to some other standard, become mysterious and unusable. It is much better to start with relatively straightforward materials, and let people extend these into more detail as they understand their needs better.

We do not want to create a new bureaucracy. We want to define just enough processes to let people weave system governance into their other work. System governance is not the main task, and it should not compete to structure the major processes around it (like work planning, system selection, and project management).

Because system governance is not the main task, tools for system governance need to be focussed. They should provide the structure for system governance, but not get in the way by pretending they do anything more.

I also think complexity sends the wrong signal, as if a more complicated product was somehow more perfect. But perfection does not come from the product.

In system governance, perfection comes from using general materials, and when you understand your needs better, tailoring them to your situation. Perfection comes from weaving system governance activities into your existing processes. Perfection comes from using the tools where they add value, and not being forced to use them where they do not. In every respect, a more complicated product would take us away from where we need to be.

Intellectually, I believe that what we are doing is right. I have worked around methods long enough to know that big complicated methodologies are rarely accepted. Commercially, I know the market can demand more features and functions, even when more is less. We can only tackle that by being clear about what we offer, and why we do not offer more.