?

Log in

No account? Create an account
Will Code For Food
 
[Most Recent Entries] [Calendar View] [Friends]

Below are the 20 most recent journal entries recorded in willcodeforfood's LiveJournal:

[ << Previous 20 ]
Friday, April 28th, 2006
8:06 pm
Third Normal Form.
I am going to present a concept here that should be old hat to most software engineers, but that seems to be unknown to most managers. I can't say that this is completely surprising, as I've even met computer hardware designers for whom the idea if foreign, even though it is directly relevant to their work.

The idea goes by many names: References vs. Values, Single Point of Change, and Third Normal Form, although you wouldn't be able to guess by looking any of the above up on Wikipedia, as they all couch the idea in their own idioms and Jargon with the result of making it seem complicated.

The idea itself isn't complicated though. Its very simple: important of data should operationally exist in only one spot and changing it there should change it everywhere. In practical terms, it says that if you keep two sets of data files that hold the same data for different purposes, you are asking for trouble when the data doesn't match. You may discover that your costs of producing a good have gone up, but your website and advertising departments don't know, and keep promising delivery using last months prices.

Now, ensuring that everyone has access to the parts of an integrated database system that they need, while maintaining security, auditability and data integrity is not simple, but its better than the alternative of having many separately updated database, using some sort of ad-hoc method to transfer information from one to another. Sadly though, this last method is by far the most common in businesses, both small and large.

I was recently talking to a small business owner about the fact that his inventory had a high turnover with products being changed all the time, but that his web page by which he sold items had a small set of just the 'standard' items. His problem was that he didn't want to pay a web designer every time he needed to update his site, which was at least weekly. It was not hard to convince him that it would be worth paying five to ten times as much to get a dynamic website that could talk to his inventory database so that every time he got a new shipment and marked it ready for sale, his web site would automatically adjust to sell the item.

Of course, to do so he would have to either switch to a proprietary web server, or migrate to a different database, because the two pieces of software he was currently using refused to talk to each other. This is why the current debate about the use of open standards for file formats has become a hot issue recently. If your software doesn't all talk the same language, you can't readily move your data between different programs created by different vendors.

Even if your major applications can all talk to each other, its usually much simpler in the short run just to create a new independent database when you need to perform some new task, even if the long-term consequences are chaos. I once read that during the 90's Oracle had so much financial data spread among so many different databases that reconciling and verifying it was a tremendously difficult task. What's more, it usually took more than a 3 months to determine how the company had performed in any given quarter, so that the company never knew what its financial situation was, only what it had been 6 months earlier.

In the fast-moving world of information technology, that kind of time lag between stimulus and response can be deadly for a company, so Oracle decided to do something about it. It started consolidating and integrating it's far-flung database network, and over the next several years slowly brought the time for year-end reconciliation down from over 3 months, to under one month, to under one week, to a few hours.

Now, Oracle's Chief Financial Officer can ask for the current state of the company, and get an up-to-date report in a few hours. What's more, when it comes time to close the books at year's end, it only takes a day. The amount of time, effort and money saved by the multi-year database integration program rapidly paid for its cost.

Nevertheless, its amazing how often I go into a retail store to buy something, only to discover that their web site's claim that the item is 'in stock' was not only inaccurate, but that the goes by a different code number in the store than is presented on the page. The third normal form is not the final answer to solving the problems of bad data infrastructure in the world, but its an important first step -- an a step that most companies have still yet to make.
Saturday, April 15th, 2006
9:56 am
Web Site Value
I've often been surprised by the lack of planning and follow-up that goes into the design and production of a company's web site. I was once in the running for the CTO job at a particular company and during the interview I mentioned how impressed I had been with the company web page. The current CEO mentioned with some bemusement that he had gotten a lot of positive feedback on the site, but that he had never gotten around to looking at it himself.

It was one of many danger signals at the company in question that caused me to decide to look elsewhere for a job. The fact that this CEO had never gone to his own web site spoke of a fundamental lack of understanding of the technology he would be hiring me to manage, and this was a supposedly high-tech company, to boot.

If a small company were to put out a full-page newspaper ad, or produce a 30-second commercial for radio or TV, you could be well assured the CEO would take an interest and would not only make a point of seeing the result, but would usually insist on having final approval before letting it out the door.  So why then do so few companies treat their web pages in a similar manner?

Web sites, if anything, demand more attention from upper management, not less. The reason is that radio, TV and newspapers are merely publishing media. They can only take a message and broadcast it. The web, on the other hand, is an interactive media, and constitute a form of dialogue between the company and the customer. Its global accessibility coupled with this immediacy allows it to have a more personal impact on a far wider audience, for good or for ill.

A statistic I came across recently said that the typical total cost to a company to produce a single high-quality web page was about $10,000. This amount was relatively  independent of both the company and of the web design firm handling the design. If the producing company insisted on a high-quality website, cheaper firms often needed several iterations of design and feedback before they produced something acceptable, while more expensive production companies tended to succeed with far fewer attempts.

Of course, this is for cases where the company in question is actually demanding a quality in their web pages. More often than not, company directors try to get the cheapest sites produced that they can, even going to the extent of hiring their son or daughter, because they understand HTML, and what more do you need?

These same directors would never think to use these same criteria to determine who should produce a radio, TV or newspaper ad. Somehow, they understand the impact of a radio spot and wouldn't baulk at paying $10,000 to see that it was done right, but aren't able to understand the impact of a website.

Just as there are literacy programs to teach people basic reading and writing skills, and numeracy programs to help people grasp how to use and understand numbers, we may need to create executive webrancy programs to teach basic fundamentals of what the web is, and what its impact on business and society are.

Until then, I fear we will continue to see corporate website which lack any sort of useful content.
Friday, March 10th, 2006
4:53 pm
Management, Paradigms and Critical Thinking.
I mentioned paradigms in a recent article, and how they can affect the working of a company. Considering how badly misused the term is in business today, I thought it would be prudent to talk about paradigm shifts, how they affect the governance of a company, and their relationship to critical thinking.

A simple definition of a paradigm is "a model of how the world works". We all have them. The reason we believe that flicking a light switch will cause a light to come on is because we have a paradigm that says the switch controls the flow of electricity to the light.

Not everyone shares identical paradigms, of course. A child's paradigm of the light switch may leave out the electricity part, and so would need to have it explained why the light switches "don't work" during a power failure. On the other hand, an electrician will have a much more sophisticated paradigm than the average person.

Clearly then, not all paradigms are equal. Some have a much greater explanatory power than others, and some are simply wrong. One could believe, for instance, that a light switch rang a tiny bell inside the walls which told glowing fairies to congregate in the light bulb. Turing the switch off rang the bell again and dismissed them.

Such an obviously flawed paradigm is not necessarily a disadvantage to the holder. If they never need to change a light bulb, or try to repair or move a light switch, they may never be confronted with its failures. Even if they are confronted with a failure, they will often prefer to extend it so as to handle that event, rather than abandon the model. Thus, they might explain power failures as ongoing labor relation problems with the fairies.

In fact, most people hang on very tightly to their paradigms and are reluctant to change them, even in the face of the evidence. When the overwhelming weight of the facts finally forces the old, broken model to be replaced with a new one, a paradigm shift occurs.

Under the new model, previously sensible behaviors now seem wrong and new approaches to problems become obvious. It is for this reason that when science historian James Burke produced his documentary series about paradigm shifts through history, he called it "The Day the Universe Changed". Once your paradigm for something has changed, its as if you are dealing with a whole new object, be it as small as a light switch, or as large as the universe.

To get back to how this affects corporations, lets consider the observation of birds flying south, under two different paradigms. The first will be that of the hierarchical ccontrol, while the second will be the paradigm of emergent phenomena.

When someone schooled in the hierarchical model of control which is so prevalent in business today (and which I have discussed previously), is asked to analyze the inverted "V" shape of a flock of birds, he will most often claim that there is a "leader" or "alpha" bird directing the flight (and holding the position of prominence at the tip of the V), and many subordinate birds following its lead. They would describe the not uncommonly seen phenomena of two broken 'V' shapes flying close to each other as being a case of two alpha birds fighting for dominance and trying to sway the rest of the flock to their individual agendas.

When someone like Craig Reynolds looked at the same phenomena in the development of his Boids program to simulate flocking behavior, he saw something completely different. None of the birds were 'leading' and none were 'following'. At least not in any traditional sense. Each and every bird was just doing its best to stay with the flock and expend as little energy as possible while doing so. Flying behind and to one side of another bird allowed the first to ride in the second's slipstream, and reduced overall work. Far from being the position of leadership that it appeared, the tip of the "V" was the location of the one poor bird that was working the hardest (as it has no one to follow) and would gladly give up that position in a moment. Multiple "V" formations were easily predicted by the vagaries of flock dynamics and had no particular meaning; certainly none involving battles for dominance.

Imagine then, how these two viewpoints will lead to very different strategies to affect the behavior of a bunch of birds, or for that matter, a bunch of workers. If one insists on finding and identifying the leader of a group, and working to influence that group through the leader, one may be in for a rude surprise if the apparent leader is merely an emergent phenomenon of the work they are doing. By the same token, if one ignores a true leader and tries to modify group behavior by imposing uniform new rules, one may be equally surprised at the results.

The real question here is not whether or not workers behave more according to one model or the other, but how, in general, do you choose between paradigms? The answer is critical thinking, a skill which should be taught as part of every management course, but which never is. As a result, the business world is rife with paradigms which have long outlived their usefulness, if they ever had any at all.

One example of this lack of critical thinking comes from the widespread adoption of the ISO 9000 standard Quality Management System. Whenever a company that I am thinking of working for tells me that they've gone ISO 9000, it sets off warning bells. My usual tactic is to ask what case studies the decision was based on. Did they do any before-and-after measurements in perceived quality of their output or of customer satisfaction? Did they notice a marked improvement in the services they got from other companies that had gone ISO 9000?

Its very seldom that the recruiter knows any of the answers to these questions. Its even more seldom that any of them are in the affirmative (so far, never). The fact is, that as a standard for ensuring the quality of a company's output, ISO 9000 has very little to recommend it. When I was researching it, I was unable to find any case studies that documented an improvement in quality from implementing the standard. Moreover, I did find studies in which the researchers admitted to having not been able to find a single ISO 9000 company that was willing to be analyzed to see what benefits the ISO 9000 conversion brought.

If we had a greater emphasis on critical thinking skills in upper management positions, the vast amounts of money wasted on these, and similar initiatives, could have been avoided.
Saturday, March 4th, 2006
5:58 pm
Implementing Change
In a previous article, I said that the company of the 21st century will have to be willing to continuously reinvent itself in a fundamental manner in order to survive. That is a tall order, and even when a company is willing, it may find it difficult to change how things currently work.

This is one of the many advantages that small start-up companies have over large established firms. Small companies have fewer people with a vested interest in the way the company is currently operating, and it is far easier to create and maintain a culture of continuous change and improvement if you start early on in the company's history.

That doesn't mean that its impossible for large corporations to embrace change and make it part of the company mindset, but it does mean that they need to carefully plan such a change in company behavior, rather than simply announcing it and assuming it will take hold.

One of the major things that a company needs in order to succeed in making sweeping changes is that they need buy-in at all levels of the operation. They need to be able to explain to everyone, including those whose jobs may be obsoleted by the changes, why its in their best interest.

I'm going to give an example of a company that went about it the right way: Schneider International, a large US-based trucking firm. In 1988 Schneider decided to bring an entirely new IT infrastructure for running their fleet of trucks. They were going to put a $3500 satellite transceiver in every truck, and tie all of the diverse departments of the company into the new web of information. It was an ambitious plan and it needed to be embraced by the entire company, or it could be a disaster.

One of the things the company did, was look at how everyone's job would be affected by the changes and ensured that where there were drawbacks (as perceived by those holding those jobs) they would make changes to the plan so that there would be outweighing advantages. So, here's how a number of the various levels of the company dealt with the change:

  • Truckers are an independent lot, and there was serious concern that putting satellite tracking into each truck would give the truckers the impression that they weren't trusted. So, the transceivers were set up so that they could be used to replace most of the paperwork that the truckers were forced to do. Log books were now kept automatically, and a trucker could send a simple message that a load was delivered and the software could handle the rest. In addition, truckers now had email to their trucks so that they could keep in touch with loved ones and find out immediately if there was a family emergency, which proved to be a very valued resource. Finally, since the location and movements of the trucks were known with far greater precision than ever before, the truckers could now request a schedule that gave them days off so that they could attend birthday celebrations, anniversaries and other important events. In the end the transceivers were enthusiastically adopted.
  • The Service Team Leaders that served as the first line of management above the truckers had their jobs change in major ways. Before, each Team Leader had been in charge of 25 drivers and mainly dealt with things like monitoring driver performance and making sure they picked up their loads on time. After digitization, most of the paperwork went away, and the Team Leaders were able to service 40 drivers each. Instead of being overseers, they became troubleshooters for the drivers. With the new technology they could contact each of their drivers each day and make sure things were going smoothly, and deal with any problems as they came up. Overall job satisfaction improved as a result.
  • Customer Service Representatives (CSRs) used to have a boring and thankless job where they dealt with many routine enquiries about the status of a load, and would have to make frustrating attempts to contact drivers. Now, most such routine enquiries were handled on the web page, and those that still came in could be easily handled as the CSRs now had the information at hand. Instead of a boring job, they now mostly got to deal with customer problems that were out of the ordinary, and could see themselves as providers of an important service. Again, job satisfaction was improved.
  • Transportation Planners, or Dispatchers, used to spend all of their time working on scenario after scenario for getting a load from one place to another in a timely manner. They had to produce multiple scenarios because they never knew if a driver would be in the right place at the right time to pick up a load, or if they could get in contact soon enough if they were. Thus they needed 3 or 4 fallback plans to ensure they could deliver a load. After computerization of their jobs, the Dispatchers now were put in charge of overseeing and monitoring the decisions made by the automated dispatch system. Rather than being replaced, the unique talents of the human dispatchers were recognized and they were given the authority to intervene in any transport decision made by the system, and change it as necessary. This has proved to be necessary only about 20% of the time, but the results in reliability have been so far beyond expectation that Schneider now issues guarantees on load delivery times, something that very few other companies can do.
So, you see, in each case a decision was made to make the change an improvement for each job that it would impact. This resulted in the company as a whole accepting the radical change to the old ways of doing things, and make the corporate culture just a bit more willing to embrace the next change when it comes, as it inevitably must.
Friday, February 24th, 2006
9:36 pm
The Value of Service.
There's a scene in the classic holiday special "Miracle on 34th Street" (the 1959 edition, I believe), in which a Macy's manager discovers that their store Santa is happily directing customers to competing department stores whenever they have lower prices or desirable items that Macy's doesn't have in stock. Shocked and outraged by this act of betrayal, the manager sets things in motion to fire the Santa, only to discover that news of this new 'policy' has spread and it has caused unexpectedly high sales volumes. It seemed that since the customers knew that if they went to Macy's first, they would get directed to wherever the best price was to be found, it became the main starting point for everyone's shopping trips.

Of course, this is just a piece of Hollywood fiction. Sending folks to the competition whenever they were better able to serve the customer's needs wouldn't really pay off like that, would it?

Well, while doing some research today on companies listed in MIT's Interesting Organization's Database (IODB), I came across an account of Schneider National, a company which I had never previously heard of, but which I may well write several articles about because in many ways they seem to understand how to succeed in the 21st century.

On the surface, Schneider National is a simple trucking company. They have huge fleets of 18-wheelers and they ship goods from point A to point B in order to make a profit. Its not an unusual business to be in, and Schneider National has many competitors yet they have tenaciously held on to their position as the largest "truck load" transportation company in the US.

Part of the reason is that they done so well is that they have embraced several Miracle-on-34th-Street style customer service polices to the extent that they have grown into whole new sub-companies. Among these are:

  • Schneider Finance: a company that helps customers finance and buy their own trucks, so they won't be reliant upon Schneider National to ship things. They even go so far as to pass on the volume discounts that Schneider National gets when buying trucks, so the customer gets rock bottom prices.
  • Schneider Intermodal: They specialize in figuring out when a customer's needs can better be met by shipping via rail or water, and arranging for the transportation in a way that is transparent to the customer. The customer pays Schneider, and Schneider pays the other carrier.
  • Schneider Brokerage: Arranges contracts with competing trucking companies to carry loads that Schneider is either unwilling or unable to handle itself.
  • Schneider Logistics: This company is essentially the entire Schneider shipping know-how in an outsourced package. So, if a company wants to manage its shipping as well as Schneider does, they can. The company will also happily analyze business models and provide an unbiased recommendation of the best shipping company to go to, for any particular need.
Looking at the above list, most company manager's would conclude that the Board of Directors at Schneider were insane. They are systematically giving away or selling many of the things that modern companies consider their competitive advantages. Most companies would never pass work off to a competitor, even if it were better for the customer, and they would never consider letting other companies have access to hard won logistical information.

So, how has this affected Schneider's bottom line? Well, it has cost them some business that they would otherwise have had, but the gains have far outweighed the losses. Because so many trucks are being bought through Schneider Finance, the company now gets even better bulk discounts on its own purchases than it did before, and that has saved the company substantial amounts. With Schneider Intermodal and Schneider Brokerage working to ensure that their customers get the best possible deal, even if it requires subcontracting to someone else, they have greatly increased customer satisfaction and have become the transport company of choice for a number of large firms. The demand for Schneider Logistics services have been so high that that company is now the second largest in the group.

In fact, the non-traditional transport services that Schneider inaugurated are now bringing in twice the revenue of the traditional trucking arm of the organization. This will probably prove key to the survival of the company in the 21st century as a variety of microfabrication technologies come on line and ultimately reduce the requirements for shipping large quantities of bulk goods across the country. When this happens, transport companies that are still holding on to 20th century ways of doing things are unlikely to fare as well.
Friday, February 17th, 2006
5:31 pm
Cowboys, Commanders and Agents.
According to Kurzweil, the current rate at which new paradigms are adopted in North America is doubling every ten years. Another way of saying the same thing is that the time a radically new way of looking at the world needs to take hold is halving every decade.

So what, exactly, is a paradigm? The word is much-misused word in business circles these days. Far too many things are referred to as 'new paradigms' or 'a paradigm shift' when they are nothing of the kind. A paradigm is a way of looking at the world, and a true paradigm shift changes everything, because once your perception of a situation changes, how you approach that situation changes as well. What once appeared to be useful ways of interacting with something or doing business now seem ill-advised and counterproductive. Reality hasn't changed, but a new paradigm makes people act as if it had.

One place where paradigm shifts have happened thick and fast in the last century is in communications. It took about 50 years for the mass adoption of the telephone to permeate businesses and completely change how people worked. It took far less time for the PC and the web to change business practices. By the time cell phone was introduced, it took less than 10 years to go from a novel invention to a ubiquitous business device, and these kinds of business changes are becoming more and more common, and closer together.

As each new paradigm in communications has rippled through the business world, the cost of communication and coordination has steadily declined. This has had a profound effect on the most economic way of doing business.

When time lags and costs of communication were high the most economic way of doing business was through decentralized, independant operators who would go on site, use their best judgements, and report back the results. This has often been referred to as the Cowboy way of doing business. You get your best gunfighter, send them in, and hope they clean up the town. If they suceed, all well and good. If they fail, all you can do is send someone else. Despite the fact that it has long been uneconomic to work this way, there are still companies that operate in this mode.

As time lags and communication costs decreased, it became more practical to use centralized decision makers who could have a broader perspective on global trends. This organizational structure with a central Commander and a far-flung hierarchy of sub-commanders and workers is still the dominant way most businesses are constituted today but it too is an outdated model.

At the present time, with massive amounts of information available we paradoxically find that it is the man in the trenches who is now in the best position to know what is going on, and to make decisions. Not only can he have access to the full information retrieval and processing that used to be the sole purview of the commander, but he also has access to all of the local information which is currently too costly to measure, quantify and report to headquarters As a result, the current most economically sound way of organizing companies in most sectors is with decentralized, fully connected Agents, each empowered by the corporation to make binding decisions and act on its behalf.

This is where the term empowerment in management originated, and is what the management gurus have been calling for, for some time now. What has so far been offered has been so watered down as to make the term, as it is currently used, almost meaningless.

In a similar manner, the phrase that 'employees are our most valuable asset' has been bandied about, but far too few companies have adopted this idea into their corporate culture, or acted upon it in any real way. The reason that the employees are the most valuable asset of a company is because they are potentially in the best position to make decisions about the future of the company. Any corporation that refuses to understand and embrace this will inevitably fall victim to a newer, more nimble and progressive company that will out-compete them.

But just because a company has embraced the idea of empowerment and the true value of their employees does not mean it can rest on its laurels. New communications technologies are continually coming down the pipe, and new business paradigms will accompany them.

Traditionally, the way that a new paradigm gets into the upper levels of management is by waiting for the old guard to retire, and newer managers, already steeped in the new paradigm to be put in their places. This is a luxury that no manager can afford any longer. The company of the 21st century is going to have to be willing to continually reinvent itself in a fundamental manner, or it will never survive.
Saturday, February 11th, 2006
11:21 am
Terminating Relationships.
Here's a simple mental exercise. Assume that your company has had a satisfactory, although not spectacular, ongoing relationship with a particular firm for several years. You have contracted out certain sensitive operations to them, and have had no reason to regret doing so.

When it becomes time to renew their contract, they inform your company that due to a change in policy after a re-examination of their core values, they've decided to shift their focus to another sector of the marketplace and no longer wish to offer their services. Should your company:
  1. Throw money at them in the hopes of changing their minds, while flailing around for a replacement,
  2. Express polite regret at the decision, tell them to call if they change their minds, and start proceedings to wrap up any current business with them, while looking for a replacement; or
  3. Immediately sever all ties, suspend all ongoing projects, change all shared passwords, revoke all passcards and issue a stern lawyers letter than should they ever even think to use any of the sensitive information they have been privy to, they will be sued into the ground?
Very few managers would ever suggest that anything other that response number 2 should be appropriate. Response 1 indicates that the company has become dependant upon this contractor for some set of services that they are incapable of performing by other means, and that there is no contingency plan in place to deal with an unexpected cessation of service. Response number 3 indicates a state of overwhelming corporate paranoia and general ill-will as well as the apparent assumption that given half a chance a company that has so far faithfully fulfilled all legal and contractual obligations would not hesitate to stab their erstwhile benefactors in the back.

Given this, why is it that I know of many companies that have made it a policy to use response 3 whenever the contractor is a full-time employee? It seems that when they are dealing with relatively powerless individuals, rather than a powerful corporation, they prefer to throw their weight around and treat someone as a potential criminal just because they have tendered their resignation.

I am not talking about someone who was fired with cause, or whom the company has been trying to get rid of due to one reason or another, where there might be some legitimate cause for concern, but about an employee who has worked acceptably for some number of years, and now wishes to move on.

This seems to me to be the height of folly. When I've been in a position to ask about this kind of policy, the usual answer I'm given is that its a standard bit of risk avoidance. Of course most professionals with sensitive information would never stoop to use it against the company, but a small percentage would. The company would prefer to lose the opportunity to have the outgoing employee train their replacement and to endure the chaos and missed deadlines from having an important person suddenly leave, than to take the chance that they might do something harmful during their last two weeks.

Even if the expected harm from treating all leaving employees as trusted professionals (which they have been up until then) was greater than the expected harm from suddenly treating them as criminals -- something I don't believe for a moment, for reasons I'll give below -- one would expect this policy to be applied much more stringently to other corporations. After all, if a single (perhaps disgruntled) employee is a threat, then surely a group of (perhaps disgruntled) consultants is a much bigger threat. They have safety in numbers, if nothing else. If one of them decides to use inside knowledge against the company, how will they be singled out?

No, it seems to me that this is just another manifestation of the all-too-common problem of the management of a company not trusting its employees unless they are in a position to be intimidated. After all, any employee who was minded to perform some sabotage or take sensitive information with him when he leaves would have long since done that before tendering their resignation if they had any idea of the company's termination policy. I, for one, make sure to back up all personal information (such as personal email) and take it home before I tender my resignation.

Even if a would-be saboteur were somehow taken unawares the first time they left a company, they would know better the second time they left one, and these days most employees work for many companies during their careers. So it would seem that this policy has only two outcomes: it catches the inexperienced or stupid criminal, and it annoys loyal workers by treating them as if they were suddenly untrustworthy.

This last point is more worrisome than it sounds. People tend to act to fulfil people's expectations of them. If someone has already been treated as untrustworthy, they will probably be more inclined to act that way than they otherwise would be. This can be dangerous for the company they have left, as its extremely difficult to remove all of the possible threat of a leaving employee. Its hard to know exactly which passwords they had access to, or which insider information they have carried out in their heads.

Why just last month I stumbled upon an encrypted link into the heart of a company I worked for a year or so ago. I had set it up (with permission) in order to work through the weekend to meet a deadline. When I left the company and they invoked policy number 3, both they and I had forgotten about the link, and it was never taken down or removed.

Had I been of a less ethical bent than I am, I could easily have gotten into their system and wiped everything clean as the link gave root access to 90% of their network. Not that this would have been the 'incalculable damage' that is so often reported by the press in response to one of these instances. Their whole network is backed up nightly, so at best I would cause them a day or two delay before they were back up. Instead I wiped out the link.

So, they have a policy of treating people as dangerous criminals despite the fact that the worst damage I could do is cause them a few days outage even with root access to their systems, and I still had this ability despite their treating me as a dangerous criminal. One has to wonder what kind of risk analysis they were running, or if it was once again a simple case of acting out of general distrust, and not stopping to think about the consequences.
Monday, October 24th, 2005
8:53 pm
Rewards and Punishment.
Note: This article was originally posted by mistake to my technaut blog instead of this one, so I am taking this opportunity to correct the problem. My apologies to those who've already seen it.

Many managers think that a main part of their task is to motivate their employees. I think it would be more productive to see their role as removing demotivating influences. Unfortunately, one of the main demotivating factors in many organizations is the action of the managers themselves. If one looks up Bad Boss on Wikipedia, one finds a litany of sins that bad managers are often guilty of.

There are two mistakes that I would like to talk about that aren't even mentioned on the list cited above, and they both have to do with the administration of appropriate awards and punishments. The first is to assume you know what someone will consider a reward or punishment, and the second is a to provide the wrong kind of reinforcement.

It is a common mistake to assume that someone will consider something a reward because it is something the manager would like. There are few folks who actively dislike praise, but a shy employee may dislike the limelight sufficiently that dragging them to the middle of the office floor and proclaiming them an excellent worker might be traumatic enough that they silently decide to never do as good a job again. Similarly there are those who are best rewarded with cash, others with tickets to a first-run movie or with added flexibility in their schedule. People differ and assuming that what pleases one person will please everyone is a recipe for disaster, and so a manager must make a point of getting to know his staff so as to avoid that sort of mistake.

I have actually experienced the directors of a company trying to put all of the employees through a hazing ritual reminiscent of a college fraternity. The directors may have had fond memories of fraternity days from college, but this was a software house full of geeks who had grown up hating the college fraternities and everything they stood for. Not only did these directors manage to remove all company morale, but it was only the quick actions of some of the upper management who could see what was happening that averted having half the staff quit.

There is also the question of punishment: the stick to go with the carrot. Study after study has shown that punishing employees is almost always counterproductive. The furthest you should ever need to go in that direction is to let someone know that they are underperforming. Even that can be avoided if there are mutually agreed upon personal goals that the employees are trying to attain. In that case, when goals and consequences of failure are explicitly negotiated up front, failure to meet a goal is obvious to all.

Why then, does it seem that so many managers still prefer to harp on failures and ignore successes? One study of managers that I saw seems to implicate what is known in statistics as the tendency toward the mean. This simply states that average events are more common that exceptional ones. Managers had noticed (some subconsciously) that whenever they praised an exceptionally good performance, the next performance was normally lower, but when they castigated an exceptionally poor result, the next was almost always better. This, it turns out, is a perfectly normal statistical result. Try rolling a die and yelling at it every time you roll a one, or praising it every time you roll a six. You'll see the exact same pattern. The yelling has no effect whatsoever on the die, and any belief otherwise is self-delusional.

People are not dice, of course, and where yelling and screaming doesn't bother a game cube one bit, it can have a severe detrimental effect on an employee's performance. After all, if you never get an acknowledgement for work done, and all you have to look forward to is abuse, why do any work at all?

Actual performance increase due to praising and the avoidance of pointing fingers is a far slower statistical phenomenon than the tendency to the mean, and so is not as easily noticed. It is a far more useful effect though, and diligent care to give praise and recognition where it is due can have a profound long-term effect on the efficiency of a department.
Wednesday, August 31st, 2005
3:21 pm
Patent Law
I debated with myself for some time before deciding to write an article about patent law. There has been a huge debate on the entire question of software patents, and there are similar debates going on in the biotechnology and pharmaceutical worlds. Many things have been said both for and against patents in general and the specific thorny areas I mentioned above, and it wasn't clear that I had anything unique to contribute, but then I noticed some of the arguments I have against the current patent system are notable for their absence in the debates.

I can understand why these arguments tend to be absent: they tend to be more abstruse and would appeal most to technologists such as I. The debaters are not trying to sway people in the high-technology camp, as those folks have already pondered the issues and made up their minds. No, the debaters are trying to sway public opinion and political will in their favor and calm rational discourse has little place in such debates, but that is a rant for another time.

The main argument that I have against our current patent system is that it is caught in a loop of ever-tightening repression of invention and innovation, which makes the job of a technological innovator harder every day. In fact, it has already passed the point where it is impossible to innovate and not conflict with some patent somewhere. The only things that has allowed new technology to continue being developed are legions of high-priced lawyers to smooth over problems when they arise.

The two main reasons for this, in my opinion, is the fact that a patent is not supposed to be granted for an obvious application of the current state of the art, and that software is all, at the heart, mathematical. How these two facts interact is perhaps not immediately clear, and I hope to illustrate the connection between them.

To start with, the requirement to not patent obvious things is a fairly large one. Just because no one has ever painted diagonal blue and green stripes across a web page does not mean that this is something worthy of a patent. It certainly isn't worthy of a patent so general that it not only covers diagonal stripes but all uses of alternating colors on a web page.

In my many years of having to deal with software patents, I've yet to see more than a couple that weren't blindingly obvious to those working in the field, and when an obvious advancement gets patented, it has a chilling effect. Suddenly it becomes unclear as to what might or might not be patented. Is this way of calculating a running average patented? Its hard to see how it could be, but then it's equally hard to see how the concept of drawing letters with triangles could be patented, and it is.

It becomes necessary to check everything for existing patents, even things you've done a hundred times, just in case its been patented while you weren't looking. Since this is prohibitively time consuming and expensive, you have to have some sort of plan to fall back on when you are caught in violation. The standard trick in the industry is to patent all of the incredibly obvious things you do, so that you can counter-claim a violation if anyone ever calls you to task.

What's worse, there is no guarantee that what you have just patented is not already patented by someone else. The patent office is supposed to protect against this, but it doesn't have the manpower to correctly handle even the simple and obvious industrial patents.

Software patents are an even more tricky case, because software is mathematics, and mathematics can undergo automated transformation. When one algorithm can be converted smoothly into another by a compiler doing its job, how do you deal with the fact that the two algorithms are covered by different patents? Do you license the algorithm as written, or as interpreted by your compiler?

What do you do when someone manages to prove that a dozen different patents all cover the same basic piece of (unpatentable) mathematics, in slightly different forms? These kinds of things have been rare in the past, but with increasingly sophisticated tools and analysis techniques, they will inevitably become more and more common.

I don't have a solution to the current crisis in the patent office, but it is clear to most everyone having to work with patents that the current system will have to change radically in the near future, or it will surely suffer a catastrophic collapse brought about by the combined weight of demands placed upon it. So, change in the patent system, of one kind or another, seems inevitable, and for those who've had to fight the system to try and get real work done, that change can't come soon enough.
Tuesday, August 30th, 2005
9:02 pm
In Praise of Older Workers.
Since I am now starting to get long in the tooth myself, I am becoming more and more sympathetic to the plight of older IT workers who are starting to suffer from age discrimination in the work place. Information technology is often seen as a young person's game, so its not very surprising that as a worker passes 35 they begin to find it harder and harder to get a job. I want to examine that perception by looking at how older workers differ from their younger counterparts and the various pros and cons of hiring them.

The first obvious advantage to hiring an older worker is the years of experience they can bring to the job. This often allows them to avoid the mistakes that younger and greener workers would make on the same tasks. Even when they have not worked in the exact same area of technology where they currently are, they will often have seen analogous situations from which they can draw useful insights.

The insights of the older worker can also help the team as a whole, as there is a natural tendency in the human species to pay more attention to advice when it comes from a from someone well-seasoned in that area, than from someone of the same age group. This can be especially true when they are better able to predict the future course of a project, having experienced many more similar ones in the past.

Although mental plasticity does decline with age, and it is harder for an older worker to retain and integrate novel new skills compared with his younger counterparts, this is at least partially offset by the fact that fewer things are very novel. By the time someone is learning their twelfth programming language, there is not going to be much to surprise them in its particulars.

There is also a major mental ability that studies have shown steadily increases with age, and that is complex mental planning. As the brain matures it steadily increases its ability to handle more intricate ideas, until the onset of senescence. Since one of the major skills that programmers rely upon is the ability to think about a large program in toto, this should favor the hiring of older programmers.

There are disadvantages to older workers as well, of course. They may not be as robustly healthy as their younger brethren and so may take more of their sick days. More importantly though, they will be less able (and consequently less willing) to put in the sorts of massive last-ditch efforts that high-tech companies are infamous for inflicting upon their employees.

Not only may older workers be less willing to deal with the stress of excessive overtime and the inevitable burnout that follows, they may have other reasons that make them unable to put the time in. Older workers often have families, houses, mortgages, cars and all the trappings of a more settled life, and these can all put demands on the same time that a company may wish to take as overtime.

The tendency for older workers to have greater commitments also has two other effects on their relationship to their company. The first is that they are likely to expect (and demand) higher salaries to support their families, and as acknowledgement of their skills. They will also have far less time to hone those skills during their off hours than their younger counterparts. This makes ongoing training programs a much higher priority in companies that have older workers, as they may not have any non-work opportunities to upgrade their skills as technology advances.

Finally, we come to what I have always suspected was the key motivation in many high-tech companies for their discrimination against older workers: they're harder to fool. Their experience has taught them when to suspect that their managers are being manipulative or untruthful. Older workers are also far more familiar with their rights and privileges under labor law, and have often gained the confidence to stand up for themselves in a dispute with an authority figure, and that last is something that many managers fear.
6:20 pm
CTO Desiderata
Since this journal is subtitled "The Musings of an Out-of-Work CTO", and I have been pointing out what I feel to be the major flaws in how most managers currently go about their jobs, I thought it might be an idea to write out what I thought made for a good CTO (Chief Technology Officer). To do that, it will first be necessary to discuss what a CTO is and isn't.

As executive positions go, CTO is a fairly new one, having not really gained popularity until the mid nineties. As such there is still a fair amount of confusion as to what the position entails and how it differs from CIO (Chief Information Officer), and CKO (Chief Knowledge Officer), two roles with which it is often confused and commingled. The descriptions I am about to give here are by no means definitive, as the use of these terms vary widely across the industry. On the other hand, the first rule of proper communication is to make sure you agree on what the words mean, otherwise you're just making noise.

So, for the purposes of this article, a CTO is someone in charge of the tactical and strategic thinking of the company with respect to new and existing technology. They are usually more concerned with the technology produced by the company and how new technological developments in the marketplace can affect the business of the company than they are with questions of which technologies the company should use to perform its day to day business. In a predominantly IT (information technology) company, the CTO is often also the director of research and development.

In contrast, the CIO is responsible for all of the information technology deployed within the company, from the computers and networking to the phone, fax and intercom systems, but not for the company's relationship to technology as a whole. The CIO is often the chief system administrator or their direct superior.

The CKO is the most recently created position of all, and is concerned more with the correct, efficient and ethical use of information than on the technologies that support that use. Their work often overlaps with both that of the CIO and CTO and so their duties are often rolled into those two, rather than being split out into a separate job. Of course, in a small high-tech companies all three jobs are usually performed by a single individual and his actual job title is mostly irrelevant.

It behoves me to point out at this juncture that my opinions about what makes a proficient CTO are bound to be biased by the fact that I consider myself to be quite good at what I do. It is like the old fairy tale in which the animals of the jungle tried to decide who should be king. The eagles thought the ability to fly was paramount, the lions believed the king needed to be able to roar, while the moles wanted no one that couldn't build a tunnel to rule them. So, please keep that in mind as you read further.

In any case, the first and foremost ability that a CTO needs is a good understanding of technology. After all, it's impossible to give advice on technological issues that you don't understand yourself.

This technological acumen should be informed by as wide a breadth of knowledge in engineering and the sciences as possible. Depth of knowledge is necessary to implement any given project, but breadth is necessary when trying to decide if a project is a good idea, and how it might be affected by new developments.

Since technological knowledge is a key to being useful in this position, and because technology never stands still, it is necessary for a CTO to continually update their knowledge through retraining, attending seminars and conferences, and through continuous and voracious reading on technology and the issues surrounding it.

Creativity and an ability to dream are also necessary to the job, since it takes an agile and wide-ranging mind to foresee coming changes in technology and their possible consequences. At the same time, a strong dose of common sense and an ability to separate the possible from the impossible has to temper that creativity, or it will be diverted down unproductive channels.

The ability to spot the flaws in technological plans is also a major asset. There have been many innovative plans that have failed for reasons that a CTO should have been able to spot. The fact that sending IP over power lines would cause radio interference should have come as a surprise to no one. So too the fact that the Sony Betamax failed in the market place. The Betamax was put together by engineers who had image fidelity as their major goal. The public, on the other hand, were far more concerned about how long a VCR could record unattended than anything else.

Finally, but no less important than what has come before, a CTO has to be a good executive manager. They need to know how to lead others, how to inspire loyalty and how to bring out the best efforts in others. They need to know their own weaknesses, and work to improve on or eliminate them.

Their communications skills also need to be top-notch, and it wouldn't hurt if they could prove this by pointing to a blog they keep of issues relevant to a CTO. :-)
Sunday, August 28th, 2005
11:18 pm
Transparency.
Transparency is one of those management buzz-words that get a lot of lip service in modern business, but seldom does it get any more than that. Companies are far too fond of their ability to hide or obscure information that they are not proud of.

The two main forms of transparency are radical transparency where all company decisions at all levels are made public, and financial transparency where all of the books and salaries are open for everyone to see.

The most common argument against both of these forms of transparency is the same: that a competitor can derive an advantage by examining the information about the inner workings of a company that embraces transparency. Case studies would tend to disagree with this assessment though, as it appears there is a far greater advantage to the transparent company in that it is now in a better position to understand its own actions and therefore can reduce, if not eliminate entirely, the age-old problem of the left hand of the company not knowing what the right hand is doing.

There is another reason that companies don't like the idea of transparency, but it is one that they very seldom voice: its hard to hide incompetence, bad ethics and illegal behavior when everything is out in the open. One extremely common example of illegal financial behavior is in the area of stock price manipulation. This is an almost universal problem in modern business, and it has only recently come under attack.

Two ways that a company can play with its stock price is by deliberately underestimating projected earnings for the coming year, so that it can later claim that they exceeded expectations, and by distorting its earnings in a year, thus manipulating its P/E ratio. The P/E ratio is a metric commonly used in financial circles to evaluate the worth of a company's stock. The P/E ratio is the results of dividing the price per share of the company by its earnings per share (the total revenue of the company divided by its outstanding shares). If a company can inflate the amount that it earned during a given time interval, it can change its P/E ratio and make its stock more desirable, which will influence its price.

By manipulating stock prices, it is possible to convince the stockholders that the board of directors of the company are doing their jobs well, and deserve their enormous salaries. Salaries that in many cases are pegged to the performance of the company's stock though various incentive programs like stock options.

While stock price manipulation may be profitable to the executives that perpetrate it, it is often extremely detrimental to the company as a whole. When a company has its earnings inflated it ends up paying higher taxes, and becomes subject to fines when the fraud is eventually uncovered. What's worse, a company that cooks its own books often doesn't know when its in bad financial shape. A company that knows it has overextended itself can take any number of cost-cutting measures from layoffs to selling off of unproductive assets. When the books have been tampered with, these actions are not taken and a company can sink ever deeper into debt without even knowing its in danger.

Similar arguments exist for the harm that can me done by hidden misconduct inside a company. When a company manager can hide incompetent, unethical or illegal behavior, the performance of the company as a whole is impeded. Even where there is no malice, transparency can bring to light hidden inefficiencies in how a company operates, and encourages the spread of good ideas outside the standard company communication lines.

When a company opts for transparent internal processes, it tends to clean up its act so as to put a good face to the world. This will enhance internal efficiency and tend to increase employee satisfaction with the company.

It can also have large beneficial effects on the relationship between that company and its suppliers and clients. Suddenly, their suppliers are able to understand much better the needs of the company, and can often suggest mutually beneficial supply arrangements that may not have been obvious in a closed environment.

The customers will also benefit from knowing how a company is actually doing business. Both from the piece of mind that this can bring, and from better relationships that spring from more open communications. In a transparent company, there is no reason to forbid employees form communicating with customers or from having blogs in which they talk about daily business in the company. Instead these activities may be actively encouraged for the free advertising and grassroots advocacy they often become.

The only drawback to transparency, if you wish to call it such, is that in a transparent company the ability of its members to communicate with each other becomes of paramount importance. Thus, where traditional companies may have dress codes, a transparent company is far more likely to be concerned with the spelling and grammar abilities of its employees, sometimes to the extent of providing remedial training in these areas. After all, which will be seen more often outside the company walls, an employees dress, or their email?
Friday, August 26th, 2005
11:00 am
Democratization of the Workplace
In a previous article I talked about setting up more natural hierarchies in a firm, where the employees choose their leaders based on competence. At the time I wanted to give an example of such a company, but although I had heard of such a thing, I couldn't find any useful references.

Well, today there is an article in the Wall Street Journal about Schlumberger Ltd., a company which is doing exactly that. Now, before I give the wrong impression, this service company specializing in oil fields has not gone all-out democratic. What is has done is formed a large number of internal communities that are separate from (and in addition to) the standard corporate hierarchy.

These internal communities are intended to bring together employees of similar knowledge, skills, and interests regardless of which country or department they work in. The communities are self-organizing, with elected leaders, and have a voice in company policy. As one consultant to the company has noticed, having an arena in which they can compete for the acknowledgement and respect of their peers is far more motivating to many professionals than the possibility of gaining stock options or a corner office.

Schlumberger did not set up this system out of the goodness of its corporate heart. As a service corporation its main resource is its technical professionals and managing and motivating them had become a major problem. This small foray into letting these folks manage themselves has so far been a rousing success, and I hope to see it spread to other aspects of the company, and to other companies.

Now, what Schlumberger is doing is not exactly new or unheard of. When Peter Drucker was consulting for General Motors in 1946, he advised them to set up something similar. General Motors regarded the idea with suspicion and declined to follow this advice.

Modern companies are now slowly becoming less reluctant to do this. Novell, for example, has instituted a similar system to Schlumberger's. Other companies have taken it even further. Mondragon is run as a coop of cooperatives and has elections to determine its governing body.

Semco, a Brazilian company, has taken the matter further. At Semco everything is run in a democratic fashion, and the company strives for complete transparency. The results have been a phenomenal quadrupling of productivity and profitiblitiy over the last ten years.

Since Semco is run by its employees, and people cannot make decisions in the dark, the company has made all of its financial books publicly available for review. They have even taken the additional step of training all of the employees so that they have the skills needed to read and understand the companies financial position. Since these folks will be making important financial decisions, it only makes sense that they be given the tools needed to make an informed analysis.

There have also been many other changes in how the company is run. There is 100% flex time for everyone, for example. Each employee determines when they will be at work, and when not without restriction even if they work on the assembly line. Folks are free to change offices at will and there is no organizational chart. There is no dress code either. Even salary and benefits are left to the employees to decide; everyone knows what everyone else is earning.

For further information on this remarkable company, consult either of the fascinating books written by Richardo Semler, the man who transformed the company: Maverick: The Success Story Behind the World's Most Unusual Workplace and The Seven-Day Weekend: Changing the Way Work Works.

For those so interested, there are other books available that talk about the same ideas, such as Inventing the Organizations of the 21st Century, The Hidden Power of Social Networks: Understanding How Work Really Gets Done in Organizations and The Myth of Leadership: Creating Leaderless Organizations among others.
Tuesday, August 23rd, 2005
4:06 pm
Management Priorities
I've recently been reading about Dee Hock, the founder and former CEO of VISA (the credit card company). He has won numerous plaudits for his ability to run companies and was even inducted into the Business Hall of Fame.

In his later years he's been working on his own personal theory of company management that he calls Chaordic Theory. This is he claims, just a renaming of complex systems theory (which he thinks has a stupid name) as it is applied to management issues. Despite the really dumb name (a compounding of 'chaos' and 'order'), he does have some keen insights into the way a company should be run, and the role of a manager in that company.

One of the things he points out is that most folks have seen synergy while watching a team or group of people perform; a group of individuals somehow combining to achieve something beyond what they ever could have done as individuals. Synergy has become one of the many buzzwords that get used by business gurus until they lose all meaning, but this it is a real phenomena: the whole becoming greater than the sum of the parts. Its what the complex systems folks refer to as emergence.

Complex systems are — unsurprisingly — complex so it is difficult or impossible to predict when and how synergy will come into being. The best you can do, according to Hock, is to remove the things known to prevent synergy and hope for the best. This, he says, is the real task of a manager.

How should a manager go about fostering synergy? Well, the big secret seems to be to let folks do their work and let them gel as a team, with as little outside interference as possible. This means reducing the number of phone calls, emails and other urgent interruptions that disrupt folks ability to concentrate on the task at hand.

In his book The Psychology of Computer Programming, Gerald Weinberg says that the best way to measure office productivity is to compute the number of uninterupted hours per week an employee has in order to work. These are solid hours in which the phone does not ring, they do not have to talk to anyone, and nothing happens to interupt their train of thought. The ideal is to have something between half and seventy percent of the total week's work hours be uninterrupted ones. Many employees would be happy to get any at all.

So, if a manager's main business is to not bother his staff, just what exactly is he supposed to do with his time? Mr. Hock breaks down the ideal manager's time spending priorities as follows:

  1. 50% of the time is to be spent on self improvement. Most managers would be better managers if they were better people. This is mostly due to the fact that being a good manager is stressful and difficult. Learning to control your temper, see the big picture, forgive the foibles of those who work for you; none of these are easy, but all are essential to the well being of both the manager and the company. Thus, the number one priority for a manager is to work on their own integrity, honesty, ethics, empathy, and so on.
  2. 25% of the time should be spent on improving the relationship between the manager and his superiors. After all, if he is going to be looking out for his employees, he will need the cooperation of the people above him in the chain of command. Having a good relationship with his superiors give a manager the kind of flexibility he needs to do his job properly.
  3. 20% of the time should be spent on improving relationships with the manager's peers. These are all the important folk who are neither above nor below the manager in the chain of command. This includes the customers, suppliers, client companies, and managers in the same level of the hierarchy. All of these folk can have a large impact on the manager and his work. Just look at what kinds of disruptions occur when two department heads are squabbling.
  4. Finally, the remaining time, all 5% of it, should be spent on those the manager has authority over. This should be plenty of time, since all that needs to be done is to hand out new tasks when asked, see to any requests for assistance, and provide appropriate recognition and rewards for a job well done. The rest of the time the manager should stand back and let them work.
When asked how one is expected to improve ones relationships with superiors or with peers, Hock replies that it is through leadership. Not the kind of leadership that involves ordering folks around. That will hardly work with your boss. No, the kind of leadership that comes from spending 50% of your time trying to be a better person, the sort of person that others want to follow, because they want to get where you are going.

Once you gain first the respect and then the the admiration of your superiors and coworkers, then you have influence over them. Not only that, you'll end up a better and happier person in the long run, even if the short term effort to become a better person can be daunting.
Monday, August 22nd, 2005
9:30 pm
Is the Hierarchy Obsolete?
For some time now, the corporate hierarchy has been under fire for gross inefficiency and inflexibility. Books like The Peter Principal have attacked it for promoting incompetence and its death knell has been sounded by pundits many times. So, the question is, has the corporate hierarchy become obsolete, and if it has, what are the alternatives?

This very question is the heart of the management science known as Organizational Development. Over the years a large number of alternative organizational structures have been proposed and many have been tried. There are companies that have tried structuring themselves like committees, ones where there are interacting and competing functional teams, ones where everyone is a member of at least two different independent hierarchies and so on. All of these structures have both successes and failures that can be pointed to by their proponents and detractors, so it becomes difficult to analyze whether any of them are better than what we are doing now.

One thing that was made clear by these studies was that the hierarchy has one advantage that all of these other approaches lack: it is the structure that humans instinctively turn to whenever they need to work together. This inclination to form a natural pecking order seems to go all the way back to our simian roots, so it is probably a bad idea to try to fight against it. Some studies have shown that when other management structures are tried, people tend to subvert them in an effort to put some sort of hierarchy in place.

That said, there is a world of difference between the sorts of natural hierarchies that groups of strangers build, and the sorts that show up in formal organizations like corporations. Corporations tend to have many more levels of hierarchy than there is any real need for. Studies in organizational development have shown that only the world's largest multinational corporations need as many as eight levels in their organizational pyramid; most can get by with six or seven.

Another major difference is in how authority is granted. In corporations authority flows down from the CEO and is parcelled out to managers on the way down. In a natural organization, authority flows from the workers upwards and is concentrated as it goes.

At first glance this might seem like two ways of saying the same thing, but there is a profound difference here. In a natural hierarchy, it is the worker who decides they need to see the manager, and not the other way around. If that manager is unable to solve the problem the worker has, then they will seek out one who can solve it. Their old manager has just lost authority, and a new one has gained it.

When you examine the functioning of a hypothetical organization built to try and mimic this way of doing business, you notice a few things immediately. The workers spend far more of their time doing productive work, and far less of it filling in reports and going to meetings. The managers are also far less in evidence. This is because any competent worker will usually be able to go several days before running into a problem that they can't directly solve themselves.

The managers are, in turn, selected by the employees for having the skill sets needed to solve the most typical problems of that worker. At the lower levels of the organizational charts these skills may just be a greater expertise in the area of work, or a greater understanding of the priorities. At higher levels, new skill sets are typically needed. Mid-level  managers need to be adept at coordination and negotiation. At even higher levels the skills that are most often demanded are long term planning and strategic leadership.

A pilot project at the Glacier Metal Company in 1948 tested some of the ideas of having a more natural hierarchy and was wildly successful. So successful that when they were bought out in 1965, they convinced the new larger company to use the same methods. That lasted until about 1990 when the last of the old CEOs retired and the new ones decided to do things the 'old' way.

So, I think that we are stuck with some sort of hierarchical system for doing business in the 21st century, at least if we are interested in efficiency. On the other hand, that doesn't mean we have to stick with the current status quo. There are better ways of doing business being researched and put into practice, and companies that don't stay abreast of these developments will be put at a major competitive disadvantage.
Friday, August 19th, 2005
8:04 am
Don't shoot the messenger.
Its a sad fact that many companies today have a problem with with entrenched poor management practices. There are many factors that one can point out as contributing to this problem, but one I will focus on today is the simple lack of adequate feedback mechanisms.

Feedback is a necessary part of the control mechanism for any complex adaptive system. It ensures that information about the results of a decision propagate back to the source of those decisions. Without feedback, a company is blind.

Naturally, all corporations have some feedback mechanisms, or they wouldn't be able to function at all. There are financial audits, annual employee reviews, or even monthly status meetings. Unfortunately these techniques are woefully inadequate.

For it to be effective, feedback has to happen on a continuous basis, not monthly or semi-yearly. Try walking down a busy sidewalk with your eyes closed for 59 out of every 60 seconds and you'll quickly get a feel for why a constant stream of fed back information is important to the successful completion of a task.

Even when seemingly sufficient reporting mechanisms are in place, they may still fail. The quality of the information must be good, the relevance to the task at hand needs to be high, the information needs to be accepted by those in control, and it needs to actually be acted upon. Most corporation's feedback systems fail some, if not all, of these criteria.

To start with, feedback needs to be actually collected and to be complete. Many companies have a problem that is often referred to as the elephant in the corner. This is a problem that, while glaringly obvious to all concerned, is never mentioned or talked about. Often in a company the elephant is a politically charged issue and those attempting to bring it to the attention of someone in authority are likely to be attacked by those same authorities. We probably all know of examples of this. Things like the fact that the CEO's brother is not competent to manage the division he's been put in charge of, but any attempt to bring this to the attention of the CEO could get one fired.

There is also a strong tendency to shoot the messenger in many organizations. When a manager is confronted with information that something has gone wrong, they may well react with anger and frustration. If they aren't careful and end up directing that anger at the person who brought them the information, all they are doing is damaging their ability to know what is going on.

This problem is often made worse when the information reflects badly on the manager himself. It can be very tempting to take the short-term approach of suppressing information rather than the more difficult approach of accepting the situation and working to change it.

Because of these various problems in the communication of bad news, some companies have instituted a policy of appointing a random member of each meeting to be a ‘devil's advocate’. That person would have the duty for that meeting of arguing against all suggestions and pointing out flaws or problems in a project. Once there is a single dissenting voice in a meeting, it becomes psychologically much easier for others to voice any reservations they may have. The fact that the devil's advocate is chosen randomly helps to disassociate the person from the task and makes it far less likely that there will be any adverse repercussions to their speaking up.

This approach is a good first step to opening up a regular dialogue between managers and employees. It isn't a final solution though. What is needed is an atmosphere in which people can feel free to speak their minds without fear of retribution. So that information about the true state of the company can flow freely.

Even then, one still has to work to ensure that the information relevant to any given decision makes it all the way to whoever is ultimately going to make that decision, and that they give it due consideration. All the feedback in the world will fail to do any good if the loop is never closed and the collected information is never acted upon.
Thursday, August 18th, 2005
12:40 am
Reward the behavior you want.
There is a simple maxim in psychology that says that if you reward the behavior you want from people, it will be reinforced. It sounds simple and obvious, but in practice most managers don't seem to be able to successfully apply this idea. In this essay I will examine what goes wrong and see if I can't find some solutions to the problem.

To begin with, its important to understand what behavior will be reinforced by any particular set of rewards. Time and again I've seen systems put into place which end up rewarding the exact opposite of the desired behavior. Stephen Covey gives a classic example of this in his book The Seven Habits of Highly Effective People:

"Why don't your people cooperate? What is the reward for not cooperating?"

"There's no reward for not cooperating," he assured me. "The rewards are much greater if they do cooperate."

"Are they?" I asked. Behind a curtain on one wall of this man's office was a chart. On the chart were a number of racehorses all lined up on a track. Superimposed on the face of each horse was the face of one of this managers. At the end of the track was a beautiful travel poster of Bermuda, an idyllic picture of blue skies and fleecy clouds and a romantic couple walking hand in hand down a white sandy beach.

Once a week, this man would bring all his people into this office and talk cooperation. "Let's all work together. We'll all make more money if we do." Then he would pull the curtain and show them the chart. "Now which one of you is going to win the trip to Bermuda?"

Clearly, this man was not rewarding cooperation. One of his managers was going to win that (no doubt highly coveted) vacation and the others would lose. The one thing that would be fostered by that reward system was competition, back-stabbing and inter-manager rivalry. How could one expect anything else?

Later in the same book, a case is given where a retail store had problems with customer service. The sales staff didn't seem to care if they helped a customer or not. The only people who seemed to be selling well at all were the store sales managers. Upon closer investigation it became clear that the this was due to the policy of sales managers being compensated based on their personal sales records. Since they were in charge of the stores, they could arrange so that sales were recorded for them, and not for the regular staff. Since this meant that every time one of the regular staff made a sale it only compensated their boss, there was a motivation to not make sales. Once the compensation system was changed so that the store manager's compensation was based on the total sales of everyone in the store, customer service improved dramatically as did overall sales.

In both examples above, the reward that was used was monetary, but this need not always be the case. Not everyone is motivated by money. In fact, for some people it can actually be demotivational to be given a raise! (This is often caused by the raise being perceived as compensation for a soon-to-be increased workload. When the recipient is already overworked, what they desire most is time off, not extra work with extra pay.)

One of the simplest and cheapest reward systems is just to publicly acknowledge that someone has done a good job. The results that accrue from noticing when someone performs well and praising them for it can be inordinately large. That said, there is a right way and a wrong way to go about it.

When you single out someone for praise, such as the single best salesperson of the month, you are also tacitly informing the rest of the staff that they have failed to achieve this. So, you have motivated one person and demotivated everyone else. The cumulative effect is naturally negative. This happens whenever there is a single exclusive goal that only one person can achieve.

A similar thing happens when the bar for recognition is lowered so far that anyone can win praise for achievement no matter how much of a screw-up they are. This turns the entire recognition system into a joke which generates resentment at its waste of time and resources.

The thing that works is to set individual or group goals and reward everyone that meets their goals. If no one makes their goals one month, then no one gets rewarded. Similarly if there is a month when everyone makes their goals, then everyone should be rewarded.

This, however, should be a rare occurance indeed. No one is motivated by being given a goal they can achieve while asleep, just as no one is motivated by being given a goal that is impossible. The goals need to be set up so that the staff need to stretch to achieve them, and that requires individual tailoring of the goal to the employee. This will require an in-depth knowledge of that employees strengths and weaknesses, but a manager should strive to have that knowledge in any case. You should also consider enlisting the aid of those who will be trying to achieve their goals. After all, who is in a better position to know what is easy or hard for them?

Whatever reward system you put in place, and whatever the actual rewards that you hand out, listen carefully to your staff for complaints about how the system works. Are you rewarding behavior that you don't want? Have you put something in place with unexpected consequences? Never assume that just because you've thought a plan through, that it will work as desired. Human beings are complex creatures and it is very easy to find that your reward scheme has hidden assumptions that turn out to be wrong. In the end, whatever you do, you will need to carefully observe what does and doesn't work, and not be afraid to change things if need be.
Thursday, August 11th, 2005
5:28 pm
Fostering Creativity.
The ever faster changing world of technology requires ever more creative companies that are able to think on their feet, and devise clever new approaches to problems as they crop up. This has caused one of the most profound management problems facing traditional businesses today: how do you foster creativity in your workers?

The problem is profound because it cuts to the very core of what a business is, and what it believes in. A close look at the typical business organization shows that it is structured in a manner almost entirely antithetical to the ideas of creative thought and free thinking. Fostering creativity in the workplace requires a radical change to the very core values which many corporations hold dear. This is because the activities of a creative thinker — reexamining assumptions, questioning how things are done, proposing new processes — are all seen as subversive and dangerous.

As a result, although a recent poll of 500 CEOs by the American Management Association found that the majority believed that creativity and innovation was the most important factor in corporate survival in the 21st century, only six percent thought that their company was succeeding in this area. This is an unsurprising consequence of how most businesses are run. When a worker is snowed under with the demands of existing projects, a new idea appears to be a spurious request to do more work for uncertain gain. Thus there is inherent resistance.

To combat this tendency there needs to be a reconsideration of corporate policy and methods at all levels of an organization. A close examination of each rule and regulation needs to be made and questions asked about its reason for existence, its impact on corporate culture, and the consequences of relaxing or abandoning it. Among the topics that need to be addressed this way are dress codes, rules about office decoration, regimentation of break or relaxation times, office hours, and the like. I used those last examples for a good reason. All of these policies are typically implemented in an attempt to make the working environment as uniform as possible. Creativity, on the other hand, needs diversity and the feeling that differences are welcome if it is going to take hold and flourish.

And make no mistake, this has to be done on a company-wide scale if it is to do any good. It is of no use having a department of creative thinkers if the rest of a company acts in such a way as to diminish or ignore the new ideas flowing out of it. An unimplemented idea does no one any good. Besides the attempt to segregate creativity fosters the myth that it is a gift and not a skill; that those in the 'creative department' have it and no one else does. Human beings are by nature creative, and a company should maximize its return on the salary invested in each and every employee by encouraging them to use their creativity to the benefit of the company.

Other common mistaken beliefs about creativity are that it is only big things that require creative thinking, or that it is an end product rather than a process for dealing with problems. Innovative thinking needs to part of every process in a company of every size, if it is to survive in the 21st century.

What are good ways, then, to foster creativity? I have hinted at some of them above and in previous articles in this journal. Treat your employees like adults. Don't tell them how to do something, but tell them the goals and constraints that you have and let them find their own ways.

Allow for the fact that different people work best in different environments. Some folks need music to think. Some need silence. Some folks are far more productive working in a café with a laptop while others need the hustle and bustle of an active office around them. Some people would do better to stay away from the office and telecommute while others need constant interaction with coworkers to keep up their interest and attention. Some folks would prefer to come in at 6:00 am and leave at 4:00, or put in 4 10-hour work days a week and take Fridays off. Some folks would much rather show up at 2:00 pm and work until the small hours of the morning. Also remember that playing and hijinks are signs of imaginations in active use and the sense of fun that comes from innovative and creative thinking. Don't discourage them!

None of these approaches are wrong, and all of them can and should,operate together in a modern working environment. That isn't to say that there should be no rules. One persons fun is an others bothersome interruption, and the music that is necessary for person A to think clearly can make it impossible for person B. There is no one solution for these conflicts, but rather they require creative thinking from the management in how to organize things so that all these seemingly disparate goals are simultaneously met.

So, now we've come full circle. If the best way to foster creativity is through the creative rewriting of company policy, and most companies have far too little creative thinking to succeed at that, how do you get started? As is usual in the capitalist world, when there is a need, there is someone willing to fulfil it for money. In this case there are companies dedicated to helping with the transition to a creative way of operation, and universities that offer management courses in creative problem solving, to name two routes that one could take.

In fact, I should mention here that some of the statistics and the inspiration for this article were taken from a recently published interview with Fred Rozenvieg of Mindrange, a company with which I have no relationship and which I do not want to appear to be endorsing as I have no first-hand knowledge of the quality of their work. Nevertheless it would appear from the interview that Mr. Rozenvieg fully understands the scope of the problem and has a company dedicated to helping solve it.
Tuesday, August 9th, 2005
8:31 pm
Respect.
Years ago I was reading an F.A.Q. for managers (which I now cannot find) that tried to explain to them how their IT staff thinks. One of the questions surprised me by its mere existence, and showed me just how far apart the thinking of the average worker and manager can be. That question was: "Why don't they respect my position?"

The reason the question startled me so much is because it contains a number of hidden assumptions that I had never realized that managers held. These are assumptions that vary wildly from those held by the typical technologist, and it will take the rest of this article to try and explain them.

To start off with, the culture of technologists has a very strong streak of anti-authoritarianism running through it. So much so that on first reading the question borders on the nonsensical. Respect? For a position? How absurd! Respect is something you give to a person, not to a title or a position! This difference of perspective is due to the fact that what gets you respect in the geek world is how well you do your job. A position IS a job, so how can it garner respect?

In addition, no IT worker would ever have worded the question that way. If they were ever moved to wonder about such a thing, they would say: "Why don't they respect me?" The fact that a manager doesn't say that seems very telling to the technologist. It would seem to imply that the position is the only reason that they might get respect; that they have no personal qualities that would make them deserving of the respect of their staff.

Here we see one of the many contrasts between the world views of the business manager and the information worker. The first sees a managerial position as a recognition of ability and worthy of respect in its own right. The second sees the management position as a boring and uncreative job that they would never want to have thrust upon them. (Indeed, many are the cases where a technologist has been 'rewarded' with a management position, only to quit on the spot!) If there is any respect to be earned there, it is by showing everyone that the job can be performed in an efficient manner that benefits both the real workers and the company as a whole. Very few managers ever pass that test.

That brings up another difference related to how IT workers perceive their place in the company. Most IT workers are at least peripherally aware that their job represents an investment by the company. The company expects to see work out of their employees that more than compensates for the cost of having them as employees. Anything else is irrational and self-destructive behaviour on the part of the company.

This is not usually a disturbing position for a creative person to be in. After all, they are directly producing goods that the company will sell for a hefty profit. It is not hard to make an argument for their continued ability to justify their salary. However, when the IT worker performs the exact same analysis of his superior's job, he usually gets a very different result.

The manager of an IT department is usually the most highly paid individual in that department, despite the fact that they seem to do the least to improve the bottom line of the company. The only justification that an IT worker can find for the very existence of their job would be if they improved the efficiency of the department to the extent that the overhead of doing without them would cost more than the manager is being paid.

So, an IT worker expects his manager to be trying his best to make the running of the department as efficient and smooth flowing as possible. Instead, they seem to do nothing but cause interruptions and produce spurious and incomprehensible deadlines. It becomes hard to imagine how the company could run any worse without the manager than with him.

Given this perception, and the apparent contradiction that the manager both stays in his job and makes no readily apparent attempts to improve how he manages the department, the only conclusion that the IT worker can come to is that the manager is completely incompetent. Since, in the geek world, incompetence puts you as far down on the totem pole as its possible to get, is it any wonder that so few managers have the respect of their staff?

If this is the position that you find yourself in as a manager, what can you do? If you want respect, you will have to work hard to earn it. Study up on the culture of the people who are working for you. You will find it a very different world than you might think. A good place to start would be with the Jargon File which is essentially a dictionary of geek language. It also has pointers to further resources to check out.

If the department is not running as smoothly as it should (and few are) then try different things to improve matters. Get input from your workers as to what most interferes with them getting their jobs done, and work to eliminate it. If there are constraints upon what you can do or how you can operate that your staff don't know about tell them. Above all, be seen as trying to make a difference. Remember, in the geek world it isn't a crime to fail, only to not try. If your staff see that you are doing a thankless job to the best of your abilities, the respect will follow.
Sunday, August 7th, 2005
3:37 pm
100% is all there is.
One of the problems that I often see in the IT industry is that managers often don't understand their staff, and are often baffled when they try to do something to motivate them and it backfires. One of the classic mistakes is to get the IT staff into a meeting room, tell them how important the next deadline is, and ask them all to promise 110% performance in order to meet it.

This usually results in a dead silence, angry retorts, or even a mass exodus of affronted people from the room. Even when the manager finds a way to coerce folks into giving such an assurance, the result is inevitably a drop in morale and a concommittant falling off of productivity. The motivational attempt has had the exact opposite effect than what was intended. So, what went wrong?

There were probably a number of mistakes made, but the biggest was asking for the patently impossible. Now, the average person will look at a number like 110% and think of it as a metaphor for hard effort. A technologically trained individual will look at 110% and think of it as 10% more than is possible. That is a major difference.

This comes from the different worlds they live in. IT people tend to be literalists as a result from living and working with computers and electronic circuits that don't have a concept of metaphor. If you put 110% of the rated voltage through a chip, you just might fry it and the circuit its attached to. When you ask a technologist to put out 110%, he's thinking that you've just asked him to destroy his health in a misguided attempt to get something for nothing. Is resentment of such a request surprising?

Even in the cases where the IT staff can see that all the manager is asking for is a hard team effort, they will still be very unlikely to promise because what is being requested is so vague. Again, this comes from a difference in perspective.

Most IT folk, unlike most other workers I can think of, have actually put in stints of 100% effort. This usually means 36, 48 or even 72 hours of intense effort without leaving ones desk for anything other than a bathroom break. Is THAT the level of work that is being asked for? Many technologists have had periods when they sacrificed social and family life to put in 80 hour weeks for a month or two at a time. Is THAT the level of commitment that is being asked for? No one knows, because the manager is being vague.

So, if you get someone to promise to giving 110% you have either asked them to promise some vague and unknown quantity of effort, or you have asked them to lie to you. If its the former, there is resentment from having to promise something they can't live up to, because they won't know when they succeed, and in the latter case there is resentment because you've just made a liar out of someone who probably prides themselves on their honesty and integrity. In either case, the motivational attempt will fail.

So, what should the manager have done? The answer is to treat the staff like adults and give them information, not vague exhortations to extended levels of effort. The purpose of a meeting, in the eyes of engineers and software people, is to exchange needed information in as efficient a manner as possible. What the manager needs to do is tell his people what the project requires, what the deadline is, who made it, how firm it really is, what the possible consequences are for making it or not, and what resources are available. Then he should simply and humbly ask for their help in achieving it.

Now, this is not guaranteed to work. Most technologists are self-motivating, but many companies impose rules and regulations that sap motivation. If you want an important deadline to be met, then you need to prove that its important. Let the workers keep whatever hours they want, and wear whatever they want. Suspend all requirements for regular reporting or dealing with email and phone calls. In short, work to remove all obstructions that hinder the workers concentrating on what's most important.

In extreme cases, rent a bunch of hotel rooms somewhere and ship the people somewhere where there are no distractions and they can get room service if they want. Treat the people like the job is important, and they will come through for you.

On the other hand, if the deadline is artificial, or was imposed from above without any regards to how much work was involved, if its not the first, or the second, or even the thirtieth time that people have been asked to give their all without adequate compensation then all hope is lost. Don't expect to get a concerted team effort from such a situation, no matter what you do.

Your only hope, as a manager, is to repair things so the work culture changes. If you aren't in a position to do that, then you are being of no use whatsoever to your workers, and are doing yourself a disfavor by sticking in a dead-end job. In such a case, I would strongly recommend looking elsewhere for employment.
[ << Previous 20 ]
www.pooq.com   About LiveJournal.com