All the implementation methodologies in the world are useless without the ability to execute them effectively.
Unfortunately the only way to effectively understand and test a firm's ability execute an ERP implementation is to have them as your implemen- tation partner - by which time it's often too late and you end up bare-chested and hairless marching down the road to the everlasting bonfire.
All the surveys and user references and slidedecks detailing 'special' methodologies in the world are unlikely to help that much in making a solid implementation partner selection. I'm beginning to think that the only way to do this kind of thing is to have a kind of mini 'proof of ability to execute' project that sits between the ERP selection and the real implementation (i.e. after you have bought licenses and bought-in to your partners plan).
This mini project has the sole purpose of reassuring you that the partner can deliver what they say and test their (and your) processes for managing this delivery via a limited scope deliverable with very prescribed objectives and also acts as a means to reassure the partner that you are not a 'client from hell'.
In other words, a 'test drive' of the partner and their people and the customer and their people if you will. After something like this, that maybe takes a month, either party can pull out without too much loss of either time/money or shirts/hair.