Many a technology implementation has been derailed by poor project management. In this post we'll look at some common failures of technology project management.
NOTE: This is the first article where I am specifically opening the kimono a bit. This should be an internal-only distribution but, frankly, that's stupid. For a little biz like GCS we're far better off spreading the word far and wide. We're a learning organization and tomorrow is another learning opportunity. A new PM is starting and needs to gulp down all of our accumulated knowledge while hitting the ground running. The failures listed below are the three most common failures of a GCS project.
Failure #1: Communication
While hardly unique to technology PM, communication lies at the root of most project failures. With technology the communication must be more detailed, more repetitive and more direct. More details are required because the subject is so foreign to many of the recipients. More repetitive because recipients will not allocate the time required to understand the projects goals, benefits or requirements. More direct because 'I Told You So' is the only defense of a PM stuck between a successful project and an upset user.
Failure #2: Training
What percent of users attend 'mandatory' technology training in most organizations? Far less than half. And it is rarely those knowledge workers and executives most affected by the technology changes. The 'I Told You So' defense works less well in this area though. A user unable to use a system can't avoid the need to go to training. The training will have to be conducted with that user, typically at greater expense in a one-on-one method, to accomplish the project's goals.
Failure #3: Hardware Specification
The level of complexity in a modern network implementation is the only reason cloud computing may gain a foothold. Is the variety of PCI bus types (PCI, PCI-X, PCI-E, PCI-64) really required to accomodate the very limited number of peripherals attached to a server? If that's the case why is every card available in every bus type? And then we're mixing SATA and SAS, internal and external with a cool dozen different connectors?
Failure #4: Documentation
The easiest task to postpone is without a doubt the most costly. Failure to document dramatically increases the long term cost of support and maintenance. And though any documentation is better than no documentation, thorough documentation can take 2-5% of the project's hours.
The solutions to these failures? I hope they're obvious. None of these issues are too complex. If you want solutions, you'll either have to pay us for it or wait for a future blog post.