Why Creating Lasting Change Need Only Ever Take You a Few Hours… Part II
On 14th February 2022, the BBC and the UK’s national papers reported on what the BBC said may be “the most widespread miscarriage of justice in UK history.” The gist of the story is that
the UK Post Office between 2000 and 2014 prosecuted 736 branch managers—an average of one a week—because senior management trusted the information spewed out by a recently installed computer system. The faulty software made it look as though Post Office money was missing from their branches, resulting in over 700 criminal convictions for theft and false accounting.
Many hardworking, honest people were financially ruined along with the lives of countless families, shunned by their communities following the conviction. Only now, after 20 years of campaigning, during which time some of the wrongfully convicted have since died, campaigners have at last succeeded in having the cases reopened. For two decades, the Post Office and the courts took the word of a machine over the word of 736 intelligent human beings—long-serving, honest, loyal Post Office personnel, each of whom had always protested their innocence.
Before the advent of artificial intelligence we did not have artificial stupidity. Before machine learning, we had no machines with learning difficulties. As with badly behaved dogs, though, we must look rather to their owners. No one can shirk responsibility for decisions by effectively delegating them to a computer. If the 2008 financial crash taught us anything, it is this.
The question is not whether to deploy digital solutions, but precisely when and where and how, to what extent, and in what way. As the father of the modern dictionary Dr Johnson remarked 250 years ago, “A cow is a very good animal in the field; but we turn her out of a garden.” Too often our prize gardens are overrun by the cows of digital solutions. My once favourite English country inn, overwhelmingly popular in the summer because of its magnificent pub garden, in its greed for expansion paved over the garden to enlarge its car park. Unsurprisingly, it was out of business within a year.
Like a larger parking lot, digital solutions promise scale, but they can obscure or eradicate some of what made the business successful in the first place. In the late 1980s, a rapidly expanding, market-leading marketing services firm saved a fortune by hiring out-of-work actors to staff call centres fielding sales enquiries from new customers and old. But as this fundamentally was an expert-to-expert selling task the company’s sales consequently plummeted.
The same too easily happens today when companies reflexly privilege digital approaches over seeking to understand the realities of customers and employees. Regimented, low-bandwidth digital solutions may seem more efficient than high-bandwidth, personalized contact with real, responsive, knowledgeable people with bags of common sense, but may similarly throw the baby out with the bathwater. And, worse, companies may not realize what they’ve lost—and who they’ve lost—until it’s too late. Artificial Intelligence can reduce the space for human intelligence, initiative, and resourcefulness.
Consider the fact that some $4.5 trillion worth of IT solutions will be sold in 2022, and hundreds of billions of dollars or more will be spent promoting these technological wonders. You can be sure a lot less will be spent on questioning whether there may not be a better, low-tech way just here, whether the intelligence required here is after all the natural variety, or whether Big Data is a big mistake.
If we are to use artificial intelligence intelligently, just as much thought and design must go into deciding which functions we will deliberately leave to the natural intelligence of real people. As with the supermarket case at the end of Part I, it is just a question of where the control should ultimately lie. It’s one thing to use AI, quite another thing for it to use you.
The Flight into Abstraction
Instead of painstakingly arriving at simpler, more effective solutions in the first place, solutions which are often far quicker, far cheaper, and far more effective, we sloppily and complacently create unnecessary complexity because we know we can leave technology to manage it for us. We tend to turn to digital solutions just because we can.
But just because you can do something doesn’t mean you should. An old Nato training manual advises that just because it is humanly possible to push a pea up the side of a mountain with your nose does not mean that this is the most effective way of getting it there or that the exercise is worth undertaking in the first place in pursuit of your greater objective.
As in the case of the empty supermarket, the now endemic, destructive disparagement of local knowledge is in danger of becoming ever more entrenched. This is due not least to the accelerating speed of technological advance increasingly outstripping our understanding of how best to deploy it— most notably in AI.
Tapping local knowledge is key to successful business, as we all know. But our own research has revealed that it’s also key to fast and effective change. This goes far beyond the question of natural intelligence versus artificial intelligence.
Computers like numbers they can crunch; it’s basically all they do. So we try and feed them numbers. In recent years, there has accordingly been a craze for quantification. Measurement is important, indeed vital. However, we habitually tend to measure not what it is important or relevant to measure, but what it is easy and convenient to measure, or what we already know how to measure, or what we are used to measuring, or what someone upstairs told us to measure. It’s pretty arbitrary and one-size-fits-all. But be careful what you measure because by and large, you’ll get what you measure.
The fetish for quantification however, is in turn only part of a still wider, rapidly accelerating vogue for ever more abstract, convoluted systems, whether analytical or bureaucratic or both. Seven decades ago, in The Human Use of Human Beings, Norbert Wiener, a pioneering cybernetician, prophetically warned of the dangers of groups of people extending their control over others, not only “by means…of machines themselves but through... techniques as narrow and indifferent to human possibility as if they had, in fact, been conceived mechanically.” In trying to fit in with the machines that we have made in our own image, we end up remaking ourselves in the machine’s image, reducing ourselves too often to unthinking automata.
Indeed over the course of my career, the single most dramatic and consistent, all-pervasive change I’ve witnessed in business, in finance, in government, and in the third sector—and the single most destructive one—has been the ever increasing abstraction of decision makers: the flight from the personal, living, real and concrete, simple and down-to-earth, to the impersonal, remote and abstract, lifeless, convoluted, mechanized and arcane.
The Lost Art of Management
The most influential business and management book of the ‘80s and ‘90s, the 4.5m-copy bestseller In Search of Excellence by Peters and Waterman, popularized the practice of “MBWA” or “Managing by Wandering Around,” demonstrating its power. When the book was published in 1982, spreadsheets were not yet something all managers viewed and tinkered with on a computer; they were still, literally, just large pieces of paper, filled-in with a pencil and a rubber eraser, used by the denizens of the accounting department, unseen.
As late as the turn of the millennium, instead of tying themselves up in endless formal meetings and dueling emails and staring at computer screens fancying they were divining the state-of-play of the company, the best CEOs and senior managers would routinely work informally, face-to-face with their direct reports, and spend much of their time wandering freely around their organization, interacting spontaneously with employees at all levels. Often they intervened ‘informally’ but deliberately as they made their rounds.
In our own experience, CEOs can also learn to do so by meticulous design, for instance when a CEO changed the previously sloppy maintenance at all the company’s plants overnight, with a single, carefully planned, spontaneous gesture which was anything but spur-of-the-moment:
On his next visit to a major plant, as he made the rounds chatting to the engineers, the CEO spotted a machine leaking oil, leaving a small puddle on the dirty concrete floor. Dressed to the nines as always in his monogrammed shirts, French cuffs and madder silk ties, the CEO, who had originally trained as an engineer himself and had risen through the ranks, casually asked for an adjustable spanner (or monkey wrench). And as he continued chatting, he got down on one knee in his pinstriped, flannel, Savile Row suit, and tightened the offending nut, without comment, carrying on chatting amicably about various issues as if nothing were amiss.
The story spread like wildfire through the company’s facilities, far and wide. It was the last unattended leak he or anyone at any of the company’s sites would ever have to see. The culture of sloppy maintenance, which had resisted all previous corporate efforts to improve, was transformed in a heartbeat.
The Myth that Big Problems Call for Big Solutions
As in the annual leave case, the “cultural” maintenance issue had been a major, intractable and costly problem, leading to production line downtime, but the solution, the change intervention, was tiny, fast, and transformative. And again, the change stuck. In a world of substance-and-forces, cause-and-effect, the world of the old epistemology, big change requires a big intervention. But in a world of the new epistemology of form and pattern, information and communication, which is the world I firmly believe we actually live in, a tiny intervention, carefully pinpointed, will invariably suffice if we can find it. For example:
Major litigation by all 50 U.S. states was underway over sales practices violations by a huge, successful financial institution which had already admitted the error of its ways and had paid several billion dollars in damages. As part of its rehabilitation it was required by the federal government to undertake a major overhaul of its management systems and practices to convince the regulators in all 50 states that they had repaired their sales practices and had mechanisms in place that would prevent such abuses from being repeated.
After spending untold millions of dollars on major training programmes and computerized sales-management systems installed by the big management consultancy companies over a few years, the large-scale mis-selling continued unabated and the regulators were growing increasingly impatient.
Finally, one morning, a company executive decided to take a different approach. The solution he and his advisers1 co-devised by lunchtime didn’t cost a cent, required only a few phone calls that afternoon and one conversation in person a couple of days later, and was fully up and running within a week, immediately resolving the entire issue and making such mis-selling now all but impossible. The intervention took the form of a small, simple tweak to longstanding protocols—which changed everything.
The states of New York and California swiftly reviewed and blessed the amended protocols, with the remaining 48 states quickly rallying as further signatories. In nearly 30 years since the time of the intervention, any infractions have been so trivial that they have not attracted a single complaint from any state regulator, and, following the executive’s bellwether example, the blindingly simple new protocol, requiring no systems and no training at all, was made standard industry practice nationwide.
In another case,
the Board had just approved a 5-to-7-year contract with a leading management consulting company, with estimated fees in excess of a billion dollars including an annual retainer of fifty million dollars. It was to transform a corporation’s culture and operations around risk, where shortcomings had been costing the company billions of dollars a year.
The CEO, just before embarking on what was certain to be a disruptive, time-consuming exercise, asked a pair of trusted scientific advisers2 if they could design an intervention that would solve the problems and create the desired change right away—purposefully not telling them a word about the proposed 7-year, billion-dollar solution, and without giving them any background about the problem at all.
Together with the CEO, the two of them completed the intervention design that afternoon, and it was implemented the following day. It worked. Less than three weeks later the CEO cancelled the big consulting contract before it started, on the grounds that the across-the-board transformation had already been achieved, the problems completely solved, and that there was nothing more remaining to be done.
The shift in thinking and in practice that makes it possible, reliably, in virtually every single instance, to create such dramatic transformations overnight by means of very small catalytic actions is a huge topic, and one that will be covered in the pages of upcoming issues of Change.
Envoi
When you look back on your own hard-won successes in your personal life or in your career, you can often identify, with the benefit of hindsight, one key thing you did—one lunch, phone call, or tiny shift in how you did or looked at something—where, if you’d only done that one small thing at the start, you’d have been spared 90% of your struggle and would have got there much sooner.
But what if you could reliably pinpoint such catalytic actions in advance, in any situation? That is precisely what the new epistemology and the resulting shift in our approach to change makes possible. But meanwhile, once you know such things are possible, then, in the blink of an eye, that already changes everything.3
© Copyright 2022 Dr James Wilk
The moral right of the author has been asserted
From Interchange Research, working at the time in New York City
Again from Interchange Research, in New York
The author wishes to thank Ir Arie Quik for his invaluable contributions to this article, including the annual leave case