Client Focus and the Halo Effect
post # 393 — June 27, 2007 — a Client Relations post
Ed Kless, Senior Director of Sage Software, just wrote me an email asking this: â€œI have just finished Phil Rosenzweigâ€™s The Halo Effect . It casts some serious doubts on the beliefs of many people in business today. In particular he calls in question the belief that customer focus causes financial performance. This has been a key belief of mine and was reinforced by your book Practice What You Preach. I would be most interested in your seeing thoughts (if any) as a blog post.â€
Well, Ed, Rosenzweigâ€™s book is, indeed very stimulating. He doesnâ€™t actually question the conclusion (or belief) that customer focus causes financial performance; he challenges the validity, rigor and logic of the various studies that purport to have proven the link. In the language of the court system in Scotland, itâ€™s a matter of â€œcase not proven.â€
Rosenzweigâ€™s main point — the halo effect — is that it is a fatally flawed research approach to identify successful companies and then ask people (internal or external) what attributes these companies or leaders have. He argues that this approach could equally well uncover attributes that are the outcomes of successful financial performance, not the cause of it.
So, he points out, if you ask observers or employees at a financially successful company if the company is customer focused, there will be an â€œafter-the-factâ€ bias to say â€œyes,â€ whether or not the company actually was, in some more solidly measured way, actually more customer focused than others.
Rosenzweigâ€™s targets are books like â€œIn Search of Excellence,â€ â€œBuilt to Lastâ€ and â€œGood to Great.â€ He doesnâ€™t say their conclusions are wrong; he says their conclusions are not even close to being â€œprovenâ€ (contrary to what the authors say.)
My book, Practice What You Preach, is not covered by Rosenzweig (it wasnâ€™t a best-seller) but I have no doubt he would make a similar critique of my methodology. I surveyed people in 139 businesses on 74 questions and explored the statistical relationships between that opinion data (on what was and wasnâ€™t going on in their office) and the financial performance of those businesses.
I think I would argue two relative strengths of my study: By allowing the statistics to tell me which of the 74 questions had explanatory power, I allowed the data to be discriminating between which aspects of office culture was correlated and which wasnâ€™t.
If you were nit-picking, you COULD argue that all I discovered is which characteristics have a high halo effect (i.e. are given high ratings by employees when things are going well) and those that have a low halo effect. But thatâ€™s a pretty complicated argument.
Secondly, I did use a statistical methodology (structured equation modeling) — used by none of the authors Rosenzweig examines — which allows you to test for causality (and the direction of causality) not just correlation. (Which is one of his big concerns.)
So, net, net, net — I think heâ€™s written a terrific book to remind all managers to beware of quick fad conclusions, and to remind all researchers and consultants that many (if not most) relationships in business that we think we know for sure actually donâ€™t have much of a solid research backing to them.
(Which by the way, was also the subject of Pfeffer and Suttonâ€™s book: Hard Facts, Dangerous Half-Truths and Total Nonsense)