My cynical answer is: find a definition of good vs. bad that most of them believe in, and that key outsiders DON'T believe in, radicalize it a bit, and frame every important objective in terms of that good/bad framework.

It will increase the cohesion, but not necessarily get things done. Examples that work are good=agile, bad=waterfall. Or good=marketing, bad=sales. Good=long-term thinking, bad=short-term thinking. Any dichotomy can be cast into this form.

A more extreme form is to start with something everybody hates in common, and let it grow from there. Shared hates reveal a lot more and have more bonding power than shared likes/loves. Shared hates automatically create a mission, a narrative tension, a promised land vision, and a common objective to bond over through an extended period.

This is what works. I am not saying it is a good thing. For the record, I don't use this method. I try to get things without relying on group cohesion at all.

Caveat: the smarter the people the less likely it is that this will work, because smart people quickly transcend any dichotomy. They generally are hard to cohere by any method. These smart, dichtomy transcenders, unless they are also capable of operating dichotomously in an ironic mode, can never get anything done. If you have a bunch of ironic dichtonomy transcenders, they'll generally get things without the need for strong cohesion. That's the kind of group I like.

Venkat