by LoveGolf » Fri May 14, 2010 1:43 pm
OK, so my last reply was a bit short.
I like the model where first and foremost, one person or team has the authority to stop product shipment if quality is below a certain threshold (usually measured by open issues or their ability to impact customer satisfaction. Having said that, it's not easy to exercise that authority unless there is a high degree of maturity around the business processes and people involved.
Product requirements should state the desired level of product quality, the testing team should certify to that level -- between the two, most of the responsibility lies. Accountability falls to Development, who owe it to the company to minimize costs for fixing up poorly coded issues. It is the software business after all, so we all know we learn some things as we go, we have unexpected permutations of user activity and system influences, so perfection is rarely achieved. No one should tolerate substandard quality but everyone in the process is responsible for achieving the goals set out.
LG