Centralized Software Testing - some drawbacks

Continuing from our earlier post on Centralized testing, lets look at some drawbacks of this type of test group organization.

Organizing testing into a centralized group tends to promote silo behavior and creates barriers to effective collaboration between cross-functional teams; mainly with development. Testing tends to happen “later”. Development creates a piece of code and often “throws poor quality code” over the wall to testing. Rather than working together closely as partners, development and testing often tend to be pitted against each other. Formal boundaries between functional groups lead to longer defect detection and fixing cycles, affecting schedules.

The nature of shared resource pools can affect development of subject matter expertise. Given the nature of resource allocation which happens on an as-needed basis, there is not much time for testers to really specialize in a particular domain. Transient test teams could also mean that building and maintaining robust regression test suites for the various projects can be a challenge.