Learning from others' mistakes or drawing inspiration from the success of another company, is all very good. and this leads to benchmarking, which is rational, considering the ultimate performance, good or bad, is figured in comparison to others.
The real trouble with benchmarking is what Pfeffer and Sutton* call 'casual' benchmarking - a tendency to get influenced by a certain set of 'best practices' and applying them at ones own organisation - only, never to get any real success.
The Southwest - United Airlines case is one particularly relevant one. Southwest practices were copied by United Airlines, casual dressing for front office and security positions, scrapping of meals onboard, and reduced time spent by planes on the ground. Yet Southwest had much greater success than United Airlines which ultimately got grounded. Likewise, the Toyota impact on the automotive industry was nothing short of phenomenal. Soon, the American motor companies began implementing the japanese management techniques and initiatives like JIT inventory management, Statistical process control, etc. However, they couldn't quite replicate the success of Toyota.
The bottom line really is to consider the philosophy behind those seemingly great practices and not just be myopic and limit onself to the immediate benefits thereof. Southwest Airlines did not terminate a single employee during their restructuring and therefore the management initiatives were successful. Toyota still remains ahead of the US automakers in quality/design features as also on productivity. Toyota's success was not based just on its techniques, but on the mindset of total quality management and continuous improvement. Their relationship with their workers enabled it to tap into their deep knowledge.
As such, Benchmarking would yield relevant insights only if proper planning and specifics of the aspects to be benchmarked are deliberated upon by the senior management. Some aspects are better considered within the industry - salaries and hiring practices at entry level positions, incentive schemes for instance. However certain other aspects such as safety practices, employee orientation, operations related practices etc can be benchmarked across industries. These and such other perspectives need to be addressed right at the beginning and then move towards the actual exercise.
"Pfeffer and Sutton are faculty members at Stanford and have co-authored "The Knowing-Doing Gap" and "Hard Facts, Dangerous Half Truths and Total Nonsense".