Friday, June 19, 2015

More on the application of the ADVISE model (of theory vs. practice)

After developing an example of a practical application of the ADVISE model, a number of questions arise: If a very big company, that has grown considerably “worse” (less ADVISE and more RACKET), is still in spite of that worsening considered wildly successful, doesn’t it somehow point to a flaw in the theory? What good is belonging to an organization that, from the point of view of its members, is “preferable” for being adorned with the “right” features, if the verdict of the market contradicts that preference, and rewards having just the opposite ones?

I’ll note first that this is a classical conundrum for any management theory: whenever It posits a number of normative features of its subject (be it qualities of a leader, characteristics of a company’s culture or functional requirements of an information system), stubborn reality presents us with as many counterexamples as the wildest imagination can conceive of instances where leaders (or cultures, or systems) that show none of those features are (either temporarily or in the long term) generously rewarded with tons of success.  Look no further than the case of Steve Jobs. Lionized as he is after his death (a visionary! A fearless leader! A perfectionist that led his company to new heights of excellence (and market capitalization)!). But according to the biography written by Walter Isaacson (quite revealing of Jobs character that he only agreed to provide information to a biographer that had previously devoted his energies to the life of Newton), and confirmed by countless testimonies every time the subject still surfaces in social media, he was a jerk, a tyrant, a bad person, a bully, a sadist and a megalomaniacal egotist. More or less what every book on leadership tells you not to be, as “better results” are obtained when you are selfless, nice, understanding, sensitive, compassionate, inclusive, respectful of other people’s opinions and the like. Just what Jobs what not… (by the way, I’m aware there is a new biography out there, becoming Steve Jobs, by Brent Schlender and Rick Tetzeli that presents him in a quite different light, as a person that learned from the errors of his first tenure at Apple and came as a much more inclusive, integrating type of leader, no surprise that both Tim Cook and Jony Ive have praised it extravagantly).

For another funny example, yesterday I came across this piece by Steve Denning The robots are not coming maintaining with very weak arguments that “the jobless future is a myth” (well, he writes in Forbes, so you can not really expect too serious journalism from such bigoted bunch of “greed is good” paean singers)… the article is crap, but it directed me to this other priceless piece of entertainment: The havoc a silly organizational model can wreak which I wanted to comment, as I mentioned the holocracy at Zappos (somewhat dismissively) in a previous post. Really, have a good read at it, because it amazes me that any adult with a minimal experience managing anything more complex than a boy scout group can fall for this egregious example of snake oil (you can have a look to the “holocracy constitution” here: Just wow! and note that this is “release 4.1”, in which “natural language” has replaced “legalistic language”, Jesus Christ, you really couldn’t make this thing up even if you wanted!). If I were a stakeholder at Zappos I would be demanding Tony Hsieh’s head for trying to implement such monstrosity, more adequate for a college fraternity than for an actual company trying to make money. The funny thing is that Zappos is not some cutting edge technology firm trying to resolve some of the basic problems of humanity, or some multinational ensemble cryin’ out for an innovative business model to support their ultra complex operations, but a friggin’ online retailer of shoes based in Las Vegas… shoes that other people manufacture, and which they just distribute…  pretending to change the world and reinvent the corporation so they can make an additional dime from monetizing what they can learn from people tastes so as to make them spend a bit more on what they wrap their feet in.

So, back to the glaring gap between what theories of management prescribe and what the most cursory observation of actually successful companies show us. There are two possible reactions by the authors of such theories. One is to appeal to the inherent complexity of the human sciences, and declare that “all things being equal” their prescriptions would have worked, but as all other things are never indeed equal, it doesn’t matter that much that actually they didn’t. So Apple would have been even more successful if Jobs had been less of a Jerk, and Zappos would be growing even more (and not suffering a 14% attrition) if it hadn’t adopted a completely moronic organizational model… so it still applies that you shouldn’t be a jerk or try to implement organizations that any 14 years old kid could tell you are plainly dysfunctional, real experience be damned (which by the way put all such theories beyond the realm of falsifiability, as they hold no matter what happens in “the real world”). 

The other possible reaction is to go full empirical, and declare that until every possible variable has been tallied and correlated and its influence weighted in a big enough sample, no conclusion can be reached. So, for example, if argue that simpler organizations are better organizations, they would first make me clarify what I mean by “better” (can they generate a greater return on invested capital? Do they exhibit higher employee satisfaction scores in internal surveys? Do they create more shareholder value in the short term?), then how I measure that simplicity (although I proposed a fairly objective measure for that: total number of different roles divided by number of members), and then they would want me to measure both in a big enough number of organizations, spanning multiple sectors, multiple countries, and if possible, multiple years, to see if the relationship holds and there is a certain coefficient in a multiple linear regression model that can be considered significant.

Unfortunately both approaches seem to me wrongheaded. The “damn the evidence, it just sounds reasonable so it must be so” is naïve and doesn’t take into account the unavoidable confirmation bias we humans have, that make us find plausible even the most unlikely propositions, if we arrive at them gradually and each seem to reinforce some belief of us. The “don’t make me loose time with intuitions, gimme the hard data” never seems to consider enough variables or, even worse, satisfies itself with sample sizes that end up being too small, because its conclusions rarely survive the test of time, as any change in market conditions tend to unsettle all its classifications (a couple books I recommended in the third post of this series provide a nice confirmation of the vanishing validity of these methodologies: In Search of Excellence and What Really Works both identified a number of companies as exhibiting the most promising features for being resilient, solid, durable and, well… excellent, and a decade after their publication most have vanished or underperformed their competition). It is a sobering reminder of how little we understand about what helps a company thrive and to what extent it may depend on pure chance (chance being just another measure of our lack of knowledge) to read The Halo Effect by Phil Rosenzweig, where he debunks a good deal of claims by the “empiricists”.

So if neither an appeal to common sense nor a recourse to hard data gathering are entirely acceptable solutions to sustain the validity of a theory’s claims, how should we differentiate crackpot theories (like holocracy) from valid ones? Well, we just have to admit that we can’t, or at least that there is no universally valid rule, as any theory can be valid within a certain domain of application and then fail miserably within another, or after some time even in the original one. We shouldn’t think of organizational design (or leadership education, or even information system implementation) as an engineering problem, where if we can measure a certain number of environment conditions and know the universal laws that the process we are designing conforms to we can find an optimal configuration for such process (be it the optimal output of a vacuum pump or the optimal concentration of certain isotope of uranium in a fuel rod). We should rather think of those endeavors more like the creation of a work of art (or, if we prefer to be a bit humbler, the building of some instrument by an expert craftsman from somewhat faulty raw materials) where we apply some heuristic rules and the best of our experience (that may not be fully algorithmical, or even “algorithmizable”) to some given elements, with their own resistance, their own peculiarities (be it the hardness of the wood, the irregularities of the stone or the shades of the paint we are about to use) that prevent us from knowing in advance how we may best approach it, and require a certain flexibility from our part to extract the best they can provide. As I explained in this post: More on science and pseudo-science there is not (and there can not be) an agreed upon rule to differentiate the “true” statements from the “false” ones in the humanistic disciplines (and “management” can aspire to be a discipline at most, if it someday gets some organizing principles and a hierarchy of authorities to put some order in the cacophony of conflicting, contradictory claims that today vie for acceptance within its domain). That is why I maintain that “human sciences” is an oxymoron (and an unnecessary one at that, derived from the sad and untrue opinion that science –meaning experimental, empirically verifiable, science- is the only valid method of obtaining “valuable” knowledge).

So back again to the question that launched our original enquiry, does it matter that “Emphast” is more successful (has a much bigger market capitalization and still enjoys a substantial lead in market share within its industry) even if it is further from ADVISE than it was in its time of struggles? Well, I still say it does matter, and I still use it as an exemplary case for a reason. I firmly believe an ADVISE organization is better in an absolute sense for its members (as any minimally conscious individual would tell you she prefers to exercise her freedom, be treated fairly and equitably, fill her time meaningfully by her own decision and collaborate with other similarly minded people as spontaneously as possible). This kind of organization, in the economic sphere, fosters happier employees, and as everybody knows happier employees tend to be more productive, make less errors and produce overall more value than unhappy ones (it could be argued if, “not all other things being equal”, the potential extra cost of keeping those employees happy was more than compensated by the additional benefits derived from their enhanced productivity and commitment, but that is, as I have already said, a question that admits of no universal answer, and which should then be analyzed separately for each particular case).

How then, do I explain the continued success of a company that was becoming more “RACKET’y” (and should thus suffer from less enthusiastic employees)? Well, with organizations, as with political systems, it is not so much how good or bad you are absolutely, but how good or bad you are compared with your competition. “Emphast” should have suffered from the degradation of its ability to give its work force what they truly wanted (starting with a viable path to partnership, or however the equivalent position was called in a publicly traded company), but for that to happen its competitors should have been able to keep offering something similar to THEIR workforces. But as it happens, the whole “IT consultancy” market (which, as much as they loved to talk about their strategy practice, and their human resources practice, and their start up incubator practice, and their M&A practice, is what they were really all about ,the rest being the icing on the cake, nice to boast about but economically insignificant) was going through a tectonic shift, from a relatively low volume, high value added orientation to just the opposite (what, using a term that few know has distinctly Marxian origins, was a more and more “commoditized” market) where competition happened mainly around price and ability to deliver at enormous scale, with the high margins of yore quickly imploding and the blazing careers of the pioneers a thing of the past. So “Emphast” could mistreat its employees as much as it wanted, because they were not going to find anywhere in the outside world the deals of old, so they would rather stay there and adjust their expectations downwards.   

And I definitely don’t need any complex study, survey, multiple linear regression analysis or even back of the envelope calculation to tell me that is what happened, that is why it happened, and that it is nicely explained by my model, which is why I find it valuable. And why I intend to keep on using it in the organizations I work for.       

No comments:

Post a Comment