Performance Evaluation in the Software Industry
Haoze Zhang is a project manager in Singapore.
The software industry is a bit of a different animal—not only because the products are virtual (with no size and weight), but also because they are so complex that it takes a long time and special skills to understand them.
Performance improvement methods have typically relied on evaluation, incentives and punishment. Basically, the system has used a closed-loop model to make sure employees achieved targets set in the previous timeline. For example: Top software companies use yearly or quarterly reviews to evaluate the performance of each employee; meanwhile, the target and suggested action plan will also be provided to target employees after evaluation. Following evaluation, firms usually perform certain actions (like a bonus or termination of contract) based on the results.
While the current methodology looks fine, I think it potentially fails with software projects.
The accuracy of existing performance evaluations
In the industrial era, we used “quantity” to measure performance. If you watched the 1936 movie Modern Times with Charlie Chaplin, the performance of a production line worker was easily measured by how many bolts he tied. However, those kinds of repetitive jobs have been replaced with automation. Line of code is one of the worst indicators for performance measurement (unless you want more redundant code to appear in your product and …
Please log in or sign up below to read the rest of the article.
ADVERTISEMENT
Published at Fri, 27 Jul 2018 04:00:00 +0000
