Human beings develop and test software. The SHAPE forum (Software as a Human
Activity Practiced Effectively) emphasizes that fact. In this forum, the subject of technical reviews often comes up. On a recent thread, the question was a rather simple sounding one: 

"Where do moderators come from?"

But the answers varied all over the place. Since we can benefit from learning what others are doing, let's take a look at some of the responses that were offered.

Respondent A says, "At our organization, moderators usually come from either Quality Assurance or the development organizations. In either case, the funds to support moderators' time come directly from the project."

Respondent B says, "We are careful to select 'strong moderators.' The moderator is expected to have domain knowledge in all other inspections. There was also a natural selection process over three to four years. Projects paid for moderator time."

Respondent C says, "In our development organization, several years ago, moderators were project members selected for their capability to be moderators (reasonableness, technical competence, ability to run a good meeting, etc.). The cost of inspections, including cost for moderator time, was a project expense along with all the other development activities."

Respondent D says, "When one of our divisions first starts to use formal inspections, the moderators usually come from the software quality group. The QA group has the responsibility to help the R&D projects with process improvement...After the projects start to do inspections the volume is then usually too high for the limited number of QA engineers to moderate, and then the projects train additional moderators. We do try to have some project engineers trained from the beginning as moderators, but sometimes we don't get any.

"In either case, QA or project engineer, we try to select someone with the following skills:

1. Must be willing to moderate
2. Must have (or be willing to develop) facilitation skills for both meetings and situations
3. Must have (or be willing to develop) planning skills
4. hould have good people skills
5. Must have credibility with the engineers--the team must allow the moderator to facilitate (note: we didn't say "technically knowledgeable of the project domain," we said "credibility with the

engineers." We like to get moderators from other projects. That way, the moderator will tend to moderate instead of being a technical lead to the meeting and add content.)

"As far as who pays for the moderators, the projects get charged for any A engineers, so in either case the project pays for moderators.

"Our bigger problem is finding and maintaining a good Chief Moderator within each division. We have found that having a strong champion and credible person with both management and engineers as Chief Moderator within each division is critical to maintain the process."

Respondent E says, "As we have worked to transfer the Formal Inspection Technology to other divisions, we have started with moderators from the centers' support organization (sometimes the assurance group, sometimes project people). In some cases, we were able to take advantage of project people who brought experience with them from outside organizations.

"Whenever possible, when the moderator was new, we tried to provide support by sitting in on the first inspection they moderated and give feedback directly as to how the inspection went.

"The key for the moderators was experience with inspections and the ability to be objective during the inspection meeting. One of my own experiences, I moderated for a rapid development project and came in as the only person from outside the development team. Without the development background, I was wondering if I was truly needed for their inspections. After several inspections, I had to miss a meeting and a person from the development team moderated. The team found that it was very difficult for that person to be objective and decided that they really did need the outside moderator for the rest of their inspections."

Respondent F says, "[Moderators come from] anywhere you can get them! Slow moving individuals...who don't run away fast enough! Just kidding. At our place, all staff get the same training, so in theory, anyone can serve as a moderator. There is a natural selection process where some people seem to be better suited to moderate, which gets them "elected" more frequently. Of course, we have the standard caveats, such as manager vs. non-manager, not the reader or author, etc. Regarding funding for inspections, I guess I could say it comes from projects, but it's hard to tell. We don't really have hard and fast rules about how projects are funded."

So, Where Should Moderators Come From?
In my experience, these are all good answers. The general answer I would give has two parts:

1. It must depend on the individual organization.
2. It must depend on the individual person.

Part 1 says that you cannot have a general organizational answer because there are so many different structures. Trying to impose somebody else's answer on your organization usually leads to disaster. For example, some people have success with moderators from QA, but in other organizations, QA is not a respected body--and respect is absolutely essential for success as a moderator.

Respect is also the reason for part 2 of my answer because people are so variable that you must have the right skills, above all. Trying to impose a moderator with the right job title but inadequate skills will kill any review.

I very much like the list provided by respondent D, to which I would add these comments:

1. Must be willing to moderate. (This is not a job for draftees, who will turn tail and run at the first resistance they get from the reviewers. You need a kind of unreasonable, burning dedication to do this job well.)
2. Must have (or be willing to develop) facilitation skills for both meetings and situations.
3. Must have (or be willing to develop) planning skills
4. Should have good people skills (All the above skills are essential. However, I place skills second to willingness because, as respondent D indicates, you can develop skills. But don't imagine that skills develop just from experience. Training is essential, and keeps being essential. You never get good enough as a moderator. Experience is essential, too, but it takes too long and costs too much in failed reviews to learn only from experience. Moreover, you'll burn out too many promising candidates if you just throw them unprepared into the fire pit.)
5. Must have credibility with the engineers--the team must allow the moderator to facilitate. (This may seem intangible, but it's essential. Not every moderator, no matter how skilled, is the right moderator for a given review. The moderator must not only be objective, but must be seen as objective by the team. That's why I always recommend that the producing team have veto power over the moderator. That way, they start the review with the agreement that "we accept this person's authority over us in the context of this review, on the process of the review." Authority over the process is far more important than authority over the content, as respondent D suggests when he says, "We like to get moderators from other projects. That way the moderator will tend to moderate instead of being a technical lead to the meeting and add content." More often than not, authority over content is an impediment to effective moderating, because the moderator is easily sucked into technical debates.)

So consider both the structure and dynamics of your organization when looking for potential moderators. And then examine the skills, willingness, and credibility of each individual candidate you consider. These are the factors that will determine where your moderators come from and how well they perform.