Engineering Squad Structuring

Perhaps a new newbie question highlighting total ignorance, but please bear with me and school where necessary.

Say you have been tasked with establishing engineering “squads”, loosely defined as an autonomous team of individual contributors from different disciplines, capable of building something end-to-end. Assume the reason for doing this is valid. We will also assume the competence composition is agreed. Lastly, lets not fixate on the practical inter-operating model of the squads at this stage.

What I am most interested in is how the individual squads are actually pulled together from a pool of skills (by tech/domain, product, service, capability, value stream / value item etc.) in various other environments as well as how they roll up into tribes. Appreciate there may not be one best answer, so what are the options.

Some secondary questions:

– Do squads still make sense in an organisation where decisions about what gets built happens before and in isolation of discussions as to where / by whom it will be built? How do organisations that have gone this route anyway navigate the complexity?

– If these squads were intended to work on change initiatives only (ie separate to run initiatives, although possibly using some of the same individuals) is a squad structure really a worthwhile pursuit? As above, how do organisations that have gone this route anyway navigate the complexity?

Thoughts very much appreciated.