Benchmarking - Correlation is not Causality

I was reading some information from SAP and ASUG this week about their benchmarking results, and I was reminded of something my college Statistics professor drilled into us: correlation is not the same as causality. That concept is also mentioned in a very interesting book I'm (slowly) reading: 'Hard Facts, Dangerous Half-Truths And Total Nonsense: Profiting From Evidence-Based Management' by Jeffrey Pfeffer and Robert I. Sutton.

For example, one of the benchmarking results found that companies whose HR departments have more headcount devoted to HR-advising work vs transactional work have higher profit margins than those whose HR departments are focused more on transactional work. That might mean that if your HR department works on transactional efficiency to the point it can redeploy its resources towards more advisory sorts of work, that the company will earn higher profits. Or, it could mean that companies with higher profits have more to invest in projects that improve HR transactional efficiency. Do the benchmarking statistics tell us about which of those scenarios are true?

Anyone who has worked in a low profit company can appreciate the challenges of funding projects to improve their work processes. If it isn't broken, you are not going to get funds to fix it. When a company is more profitable, it is easier - relatively - to get funding for such projects.

It's also true that when HR can spend its efforts in an advisory role, human resources are managed better and that can have a positive affect on profitability.

I suppose my point is that simple statistics and graphs are indicators and starting points for digging deeper into understanding a scenario. They tell us more about the 'what' part of a situation, not so much about the 'how'.