Perhaps the most common hazard for teams is a lack of consensus. It is a particularly thorny problem, and brings a lot of teams to heated arguments, division and disaster. Bill Isaacs of the MIT Dialogue Project believes that ancient societies were highly skilled in making consensual decisions, but that we have largely lost their knowledge and ability. Corroborating this theory, Marlo Morgan, MD, has documented the extraordinary consensus-reaching abilities of the Australian aborigines. Both Isaacs and Morgan believe that we can, with training, rediscover and apply this ancient skill.
The word consensus comes to us from Latin roots meaning "shared thought". Consensus does not imply complete agreement, but does involve seeking a decision with which everyone is reasonably comfortable. To accomplish this, everyone will need a fair opportunity to be heard and latent issues must be explored to the satisfaction of the group.
Many different tools can be used to build consensus. In fact, all the tools used in quality management contribute to consensus. For example, a well-run brainstorming session can get lots of ideas out onto the table and give everyone a chance for input. Still, most groups approach a point where they must choose between options, or try to narrow a list from many items to just a few. For this, effective tools specifically for building consensus are used.
Effective tools will do the following:
For long lists of items which must be whittled down to a few contenders, multivoting is one of the quickest tools. In a multivote, each voter picks a set number (often 1/3 or ½ the total number of choices) of items from the pool. The items which are picked by the most voters stay on the active list; the others drop out. For shorter lists of choices (10 or less) other techniques come into play. Rank-ordering is used frequently, but it can be misleading since it forces intervals where there may be none and imposes equal interval ratings where they most likely don't reflect reality.
Rating scales, in which each item is rated on a scale from 1 to 5, or 1 to 6, or even as much as 1 to 10, tend to be more accurate. Wider scales encourage gaming, and may allow the votes of extremists to distort group decisions. Narrow scales, though, may not allow enough scope for voters to express their real judgments. Since voting is not intended to make a decision, but to structure discussion and thought, narrower scales that dampen extremes of opinion are often most useful.
It is often useful to figure out what criteria are most important to the final decision, and then to rate each option with respect to each criterion. For example, you could rate three different cars with respect to price, safety, cargo capacity, performance, and styling. To make this method even more powerful, each criteria can also be rated, thus weighting the item ratings each voter makes. This division of the decision into bite-size chunks can be very helpful, and will often bring out latent criteria or opinions. This multi-criteria voting is usually done via a matrix. More sophisticated decisions can be reached by comparing pairs of alternatives, and recording votes in a matrix. These matrices can be quite powerful and are not very difficult to use.
The bottom line in consensus building methods is really this: vote to reduce a list to a manageable size, not to make a decision directly; discuss to elicit the opinions of team members and to gain insight. A vote should not be the final step in the attainment of consensus; rather, a course of action should emerge, and everyone should have a chance to assent to the group's pursuit of it.
PathMaker's Consensus Builder tool uses two main methods to support consensual decision-making. One is structured discussion, in which decisions are carefully framed, alternatives systematically discussed, and notes taken. The second method involves the effective use of voting -- rating systems and multivoting -- to reduce lists and quantify opinions.