If both think it’s the other one’s fault, neither one fixes it

I worked with a new telecommunications service provider that had started to send large volumes of voice traffic to China. As part of their buildup, they installed additional equipment and software from a third-party vendor. Initial computer-automated tests showed good call quality to selected cities and so they initiated the switchover early. The launch was a total flop.

The telecommunications service provider received many complaints about calls that did not connect and they had to replace the new service with a more expensive alternative. Accusations flew. The service provider said it was the fault of the equipment and software vendor. The equipment and software vendor said it was the fault of the service provider or even their customers who were dialing incorrect phone numbers. Both pointed to the initial successful tests as a reason for why it was the fault of the other. Nothing happened.

I didn't know who was at fault. I simply made 100 phone calls using the problematic telecom system. I made several calls to every province in the country over a couple of days.

About 1/2 of the calls failed. I knew that something was wrong so I looked at the location of the calls that failed; they were all in the same set of provinces. The successful calls were in the remaining provinces. Could it be as simple as that?

I talked to both the service provider and equipment/software vendor. We eventually discovered a slight difference in the way equipment across Chinese provinces handled the calls that led to the high failure rate. A slight tweak and call completion rate almost doubled.

Problems sometimes have simple solutions; taking an open approach without assigning blame can help us solve them.

Bear Stearns and the Tour de France

There is another case of email messages being used to identify possible wrongdoers in the financial services industry (see “Prosecutors in Bear Case Zero In On Email”, Wall Street Journal).

Anyone in the US financial services industry should be well aware that regulatory compliance requires that employee email, IMs and file uploads are tracked and cached. Over the last decade when prominent cases emerged where old emails were used as evidence I thought that this common knowledge would not so much improve ethical standards as push unethical behavior to other forums, such as the phone, face-to-face meetings and the use of allusions.

So it is surprising to see people make the same mistake as recently as mid-2007, as the Bear Stearns employees’ old emails may show that they acted contrary to their clients’ interest. My take is that as long as enough money was being made, sketchy actions that improved performance did not attract notice, but once the market tanked, it became a matter of assigning responsibility. Electronic messaging supervision (the active monitoring and spot-checking of emails by compliance personnel) helps catch problematic behavior, but not even the best system can catch every ethical lapse as it occurs.

Breaking the rules in sports is similar. In ultra-competitive contests such as the Tour de France, cyclists have used doping in order to qualify and perform above their natural peak. Again, this strategy calls for breaking the rules to improve performance. Doping is not used by all competitors, but one must assume that all competitors have the ability to use it. Prize money is high. Some are caught and some not. And in the Tour de France, this has been going on for the last century.

Hence the paradox: there are industries and contests so competitive that it can be necessary to break the rules in order to compete. Governance bodies then often eventually identify and ban these same rule-breakers, but only after they have done the damage. A behavior may be prohibited (insider trading, options backdating, doping) but the means to catch the behavior often lags behind the means to evade detection.

Perhaps the only means to reduce these types of unethical behavior is to reduce the dramatic temptation that huge compensation gives to the competitors, whether in the form of a massive annual bonus or prize money and future sponsorship contracts.

Cost/Benefit analysis and the Bryant Park chairs

There are several thousand folding chairs at Bryant Park and by the front of the New York Public Library. None of the chairs are collected or locked up at night: they stay out all the time, a fact that sometimes surprises people. Does it make sense for it to be otherwise? Back of the envelope calculations follow.

Assets
Number of chairs estimate: 3,000
List price $149 per set of two; estimated volume discounted price of $50 each
Total estimated chair asset cost: $150,000

Setup time calculation
Number of chairs that can be set up / collected per person per hour: 600
Number of people setting up / collecting chairs: 5
Time to set up / collect chairs: 1 hour each

Chair setup and collection costs
People involved: 5
Hours involved: 2 per day
Wage per hour estimate: $8
Assume chairs are set up daily: 365 days per year
Total cost of chair setup and collection: $29,200 per year

Chairs replaced per year
Wear and tear: 400 chairs x $50
Total: $20,000

Chairs lost to theft: 5 – 6 per year:
Total cost: $250 - $300

You might quibble with some of the numbers, but the direction is on target. This means that for Bryant Park, there is no economic reason to have people set up and collect the chairs every day. Chairs lost to theft do not even cost 1% of the overall labor, to say nothing of lost use of the chairs at night and I’d even argue, a loss of goodwill. For businesses, trust can be a good value.

The above is representative of some of the work Inticiti does.

On the chair I sat in today, the following plaque:
“Look at all the people”
For 10 years he helped make Bryant Park great.
Mel Kopelman 1926 – 2003.

Thanks, Mel.

Keyboards, convention, efficiency and luck

When Christopher Sholes designed the QWERTY keyboard layout for early typewriters, it was partially to separate the keys of letters commonly used together and thus prevent key jams, which had a secondary effect of slowing down typists once key jams were no longer a problem.

But still, the QWERTY standard is seen on almost all English language keyboards even though it is not the optimal placement of keys. Rather than learn a new arrangement, like the Dvorak keyboard, most people prefer to stick with what they know. (The world’s record for fastest English language typing is held by Barbara Blackburn, using the Dvorak Simplified Keyboard. She had a peak speed of 212 wpm.)

So while most of the English language world uses a substandard keyboard layout what impact does this have on typists of other languages? Is it possible that some languages, purely by luck of letter frequency, use the QWERTY keyboard more efficiently than English language typists?

I looked at the languages for which I had data: English, French, German, Spanish, Esperanto (why not), Italian, Turkish and Swedish and mapped letter frequency to the QWERTY keyboard. A screen shot is below.

Doing a true comparison of keyboard efficiency in different languages calls for typing speed tests (of letter, not word speed), but lacking these data I made some assumptions. To compare the effect on different languages, I assumed that the Dvorak model was ideal and measured the degree to which the foreign languages followed that model of putting high frequency letters in the middle row and weighted to the right hand.

And the winner is… Esperanto. Yes, by chance, this artificial language should have a faster typing speed than the others. Losers in the speed contest are German and French, both of which have their highest frequency letters in the top row and to the left. That’s the power of convention and luck on efficiency, whether on keyboards or in business.

Then again, maybe we shouldn’t focus on typing fast but rather on writing thoughtfully and well…

Fool’s return?

Beatrice Otto’s book “Fools Are Everywhere” reintroduces the role of the Fool (or Jester) in history around the world. In many past royal and religious courts, Fools acted as a type of advisor to the leadership, having close access to the leadership and the context in which to say anything. Only the Fool could speak so directly, simply and humorously to the leadership and by doing so change accepted thinking and get results.

With the commonplace seriousness of many business organizations, this can a difficult role to reintroduce, though some organizations have attempted just that. A notable example was the Corporate Jester position that Paul Birch held at British Airways (until he was… dismissed). Smaller, more entrepreneurial organizations may take more kindly to the role, but then again, by being less hierarchical in the first place, they may not need it as much. David Firth, author of "The Corporate Fool" offers instruction on the roles that a Corporate Jester should take to make a positive organizational difference. These roles are: alienator, confidant, contrarian, midwife (creative thinking), jester, mapper (knowing who knows), mediator, satirist, truthseeker.

These are basically the types of behavior we would expect to be found informally throughout a healthy organization.

Would Fools have made a difference to the behavior of businesspeople engaged in dubious behavior, such as stock option backdating, sub-prime mortgage issuance, or accounting scandals? This is a tough question but I believe that the difference is that Fools in the past could absorb enough of an understanding of the issues to comment with clarity, direction and wit. Today’s business problems may be more complex (as in the financial scandals) and impenetrable (requiring technical knowledge).

Therefore, if the role of the Fool has a future in modern organizations, it must be spread out among the many, not concentrated in one person. I believe that some creative organizations have accomplished just that through selective hiring and an open culture. We’ll see where it goes.

Speed to marketer

When I first incorporated Inticiti, I was astounded at the speed with which direct marketers contacted me. I incorporated on a Thursday. The following Tuesday I received my first piece of junk mail addressed to the business.

Certainly the process for sending that piece of junk mail was automated and not specific to me. And I did what most people would do – into the recycling bin it went.

The marketers send so much mail because they believe based on their research that each letter they send is worth more in future revenue than the cost of its printing and mailing. What happens if we become a customer of one of those companies and need to call their help desk? General customer experience notes that wait times are excessive (or “feel” excessive) and staff are often not as helpful as we would like (or we “feel” that to be the case). It sometimes seems as if companies believe that each customer served is worth less in future revenue than the cost incurred to serve them.

The second complaint (helpfulness) is more difficult to improve since it requires not only good automated menus and scripts for call center agents but also a degree of familiarity and longevity at the call center, where employee turnover can be 100% per year. The first complaint (wait time) could be easily addressed: simply add more agents. Using knowledge of call busy hours and queuing theory we know that adding a handful of agents to a large call center can dramatically reduce customer wait time. And agents in general are pretty cheap.

So why do we ever need to wait on the phone for the help desk? I think it is mostly a matter of context. Companies traditionally viewed their facility and support staff as costs rather than as potential sources of revenue. Only recently are companies thinking about call centers not as reactive cost centers, but as proactive sources of additional revenue.

Like call centers, direct marketers do not consider costs that exist outside of their business model. The most important one of those externalities to me is the time that I have to take each day identifying and discarding junk mail. Further, there is the municipal cost of disposing discarded mail. And do I throw it away or recycle it? Further to that is a possible increased exposure to identity theft made possible by each additional letter with my name and address on it. Taking all of this into account, perhaps the social costs of junk mail (or call wait time) are greater than the economic benefits that accrue to the marketing organizations and their corporate clients.

Taylorism or Tailorism?

It’s been almost a century since the publication of The Principles of Scientific Management. Is there a new Frederick Winslow Taylor, a student of knowledge worker scientific management?

There are a number of drivers in knowledge work and a number of reasons why Taylor’s approach does not work well in this area. For example, there are cost savings arguments to standardization in more complex organizations. Contributing to this is lower employee tenure. According to the Bureau of Labor Statistics, average employee tenure fell since 1996 (from the BLS study “Median years of tenure with current employer for employed wage and salary workers by age and sex, selected years, 1996-2006”). Tenures decreased the most for employee ages 35-44 (a drop of 7.5% to 4.9 years), ages 45-54 (a drop of 12% to 7.3 years), and ages 55-64 (a drop of 8.8% to 9.3 years). When employees change employers more frequently, they must spend more time ramping up in order to add value in their new organization. It would seem that lower tenures add to the benefits of process standardization for knowledge workers.

But there is a limit to how deeply an organization of knowledge workers will want to standardize processes. The problem is that knowledge work tends to be less easily compartmentalized and that knowledge workers, like the manual workers of Taylor’s factories, often resent being told how to do their jobs, especially as it may make them feel like cogs.

What do the stakeholders -- the organization, the employees, the customers – gain and lose from a process-heavy approach? To the organization: possible gains in productivity but less flexibility to act on creative ideas that produce future productivity gains. To the employees: possible improvements in efficiency in some tasks but lower job satisfaction, depending on disposition. To the customers: perhaps lower prices, but possibly less customized service.

Then I thought of my old tailor from when I was based in Hong Kong. His business was incredibly customer-centric. I would stop by his shop, have a cup of tea, talk about work, look at some fabric, talk about suit styles, talk about the family and then make a purchase decision. A very long, inconsistent and inefficient process. Behind the scenes, cutting cloth and sewing, he may have been efficient, but to the customer he gave a personal and, well, tailored series of interactions. I’d say that to some degree the knowledge worker is a customer of his/her employer.

Maybe not all of what passes as inefficiency is true loss.