David Foster had an interesting piece up a couple of weeks ago on organizations, using Moltke’s refusal in August 1914 to turn around the troops on the Western Front to attack the East as an example. Moltke was adamant that the plans in place were at the time irreversible, but the German military railway expert later claimed that he could have turned things around. Whether or not the post-hoc analysis was correct, the actual expert, of course, never got to speak with the Kaiser.
This points to one of the problems of organizations as they ossify – that information gets filtered by each layer in the hierarchy as it passes up a silo. Each layer of spin holds the possibility of not so much adding perspective as simply moving the information content further from reality, and in some organizations any resemblance between actual observations and the information contained in top management briefings is purely coincidental. CW’s NoSuchBlog had a nice post up about that same phenomenon at work in the CIA:
Our bureaucracy has become so gigantic and stultified that the “one guy” is a disconnected figurehead – whether a Porter Goss or a John Negroponte – who relies on a massive staff of lesser bureaucrats who are supposed to figure things out. But the guy who really looks at all the evidence and figures stuff out is at the bottom. His opinion, rarely, if ever, bubbles to the top. And when it does, it is so filtered, or spun, as to be unrecognizable. On the rare occasions when a lowly analyst has the chance to present unvarnished opinions, those opinions are seen as belonging to the lowly analyst – not to a trusted lieutenant of the decision-maker. I won’t get into how convoluted the actual decision-making process has become.
That state of affairs is not an excuse for micromanagement, however, since the cure in that case is as bad as the disease, as David pointed out:
An executive, of course, must have confidence in his immediate subordinates, and trust them to have gotten the necessary information from their subordinates. Otherwise, it would be impossible to run anything. To continually demand information directly from people several layers down, using direct reports only as messenger boys rather than as evaluators and decision-makers, is destructive to any organization. But it is also bad to have an organizational culture in which any bypassing of the hierarchy–as in bringing von Staab directly into the conversation–is automatically viewed as undercutting someone’s authority (which is probably how Moltke would have viewed it.)
This is the conundrum that plagues large organizations everywhere. Most big organizations fall into the Kaiser’s model, given the wide-ranging demands placed on senior managers, in a process I like to call attention deficit decision-making. There are a couple of commonalities I’ve noticed in good managers over the years I’ve been involved with big business and the government.
One of the first characteristics is that good senior management has both breadth and depth. In some companies and organizations, managers work their way up their silo and then the silos fight for the right to put one of their guys in the top slot. The resulting CEOs have depth, but little breadth. In technical organizations that require matrix teams to oversee projects, this is a very bad system. On the other end of the spectrum is the organization that rotates its fast-track managers through roles so quickly that they never have time to develop a deep understanding of their subordinates’ roles. I would argue that breadth without depth is the worse situation, but “worse” is a subjective term and depends on the business and organizational context.
My ideal senior manager would be one who spent two long stretches in two different silos, with one, perhaps two, shorter stretches or secondments in other areas. In a commercial business, one of those areas would have to be sales, marketing, or after-purchase customer service, in order to ensure customer focus at the top levels. In a technical organization, one of those areas would have to be product development, in order to get an idea of timelines and physical realities involved in making an end product.
Managers with both breadth and depth have a huge advantage in attention-deficit decision-making. They have the experience and intellectual background necessary to make decisions based on less than total information, and the intuition to be able to tell cow shit from cake batter when it’s presented to them in a silver bowl.
One byproduct in the process of producing those kinds of managers is that inter-silo communications should increase. I recently participated in a senior management review in my silo where the most senior manager present spoke about a piece of information that betrayed a prior briefing from one or other of the other silos. Some people present did not appreciate that, but I am very glad that my senior management is getting information from more than one source.
Another characteristic of good management (and one that has more to do with culture than an individual’s skills) is that organizations that are as flat as possible seem to have better managers, in my experience. The Peter Principle is a real phenomenon, and many excellent people would be better off if they could remain in their current positions. One way to accomplish this is to have two tracks of employees, a technical and managerial track. A lot of companies that attempt this do not adequately incent their technical tracks. Without adequate compensation at the higher technical levels, there is still a gravitational pull towards the process that Peter described. I see nothing wrong with a manager making much less than his or her direct reports, however, and I think the system can work with adequate compensation on the technical side. Even among people moving from technical to managerial roles, I’m always in favor, where possible, of managers keeping one minor project from their previous role – it keeps them grounded and keeps them from forgetting things as they advance.
The other side effect of a flatter organization is that there are fewer informational checkpoints where information can be spun or deleted. I think a lot of the problem in modern corporations and government stems from where we are in the history of human development: an awful lot of employees are de facto parasites – if they were eliminated the organization’s productivity would only be marginally affected in the short term (I’m not even talking here about incompetents, the elimination of whom would result in an increase in productivity). No one wants to take a hard look at staffing because managerial prestige is measured by the size of the subordinate staff. The actual number of real economic actors in a large corporation, just as the actual number of combat troops in an army, is actually surprisingly small. The problem with attempting to streamline an organization is that senior management often has a very wrong idea about where the actual value gets added in their organizations. Also, value may be added in a delayed feedback loop – someone’s absence might not be felt for years, when they are needed at a critical juncture, and it is hard to measure lack of revenue from products not developed or ideas not generated by the absent employee. However, most people’s jobs put them in a semi-parasitic position at least part of the time, hence the multiplication of hierarchies and in-fighting, and the fight for control of information.
Getting the right information to senior management is much easier when senior management has enough sense to know that there are people who ought to be in the room when decisions are made (and even has an idea of who those people might be – managers should set aside time for “management by walking around”), and when the critical experts are not buried under multiple layers of middle-management.