Month: April 2010

Benchmarking Projects

With regards to benchmarking, there is a chinese saying “人比人，气死人” (You will only anger yourself when comparing yourself against others).

When it comes to benchmarking, perhaps it is most useful when comparing against similar projects within the same organisation since the baselines (company culture, company structure, etc.) should be the same. However, what is most important is that project reporting is truthful and the management is willing to accept deviations from the benchmark if there are valid reasons.

Problem is, we’re talking about corporate flexibility when it comes to standards. “Flexible standards” comes close to being an oxymoron, and few are the corporations that have management willing to commit to such an undertaking.

Paper certified, but practically uncertified

In the IT world, certification is given quite heavy weightage when it comes to beefing up your resume, however employers should be wary that a certification just means that the person knows about the best practises behind the certification standards but does not necessarily mean that they will make use of those skills.

For instance, I have come across PMP and PRINCE 2 certified professionals who have detracted far from what is preached by those standards. Some of them may have been restricted in the sense that they need to fit into the corporate culture, but others may just lack the necessary skills (both hard and soft skills) and are just hiding behind those pieces of paper.

Certification may be important, but it is not what defines the person. Ultimately, since IT work usually involves working closely together in teams, such cert-rich but skills-poor people will get exposed and lose their credibility.

Using Test Driven Development

Although I like the concept of Test Driven Development (TDD), I have yet to apply it in a corporate context as it might be excessively off the beaten path for management to accept comfortably.

As it will be difficult to explain TDD within 3 paragraphs, I have left it to Wikipedia to do the honours. Basically what it entails is to write the test case first before writing the code that is expected to pass the test case.