We suggest an approach to testing methods in an industrial setting without control groups or much of any rigor other than long-term success in the market.
Most books on agile methods are written by folks who tried something and found it worked. Books about methods that don't work don't sell. The net is a lot of books that explain methods that are hard to duplicate and don't yield expected results even when performed as described.
We will describe processes that vary in small ways from existing work practices. We will rate these proposed processes by established criteria where more alignment with a specific criteria suggests a specific result. Longitudinal review that yields expected results will be said to have succeeded.
See Criteria of Interest to be considered.
See A Switchyard of Methods covering all criteria
See Moment of Creative Potential not gone to waste.