Who to involve in a design review

The main rule of thumb is to involve representatives of all the major stakeholders in the design, but to keep the design review team to a manageable level. Organisations where design reviews are part of a formalised development process will have existing guidelines for who should be involved in a review.

In most reviews, the majority of reviewers will be busy engineers who have “their own work to do”, and don't want to be unduly burdened by reviewing too many other engineers' designs. Establishing a culture where review is a normal part of the design process initiated by every engineer at the point where sufficient work has been done for constructive feedback to be gained helps everyone to see reviewing other people's work as part of their own work. Having too many reviewers can bog the process down though, and may lead some reviewers to assume that someone else will look at the details, so they don't have to. Well-partitioned designs need fewer reviewers than designs with complex interfaces and lots of coupling between components. More than about 5 or 6 reviewers starts to get unwieldy, except for key system components and interfaces with many stakeholders.

It is often valuable, and sometimes mandatory, to include an independent peer in the review process. Their role is to provide objective feedback as someone divorced from the design of any of the system. They need an understanding of the technology involved, some domain knowledge and the ability to comprehend and pick up new designs quickly, since this may be their first exposure to the design. However, they need not be an expert in the area in question; in fact, it can be helpful if they have relatively little knowledge in the area of the design. If a smart engineer with little prior knowledge cannot comprehend how a design works on the basis of the design materials under review, it suggests that the design is lacking in description or perhaps incomplete.

---

Social Bookmark this page: