Computers operate on binary logic – one or zero, true or false, yes or no.

I guess it’s contagious.

A while ago I gave a simple piece of advice: Never say no. Quite a few correspondents ignored the essay that followed. Instead they applied the binary system to my advice, inferring that I’d told them to always say yes.

False inference. The binary system is wonderful in its place. Applied to human decision-making, though, it’s usually disastrous, forcing a universe of possibilities into two oversimplifying buckets.

To illustrate, imagine a business manager asks you to install a Windows NT server for his department (we’ll pick on NT today – next time we’ll choose a different OS to victimize) so he can run a particular packaged application, in violation of your standards and personal judgment.

What do you do?

You can say “I’m sorry, but that would violate our standards,” or, “We try to avoid NT because it’s far less stable than the other solutions on the market.” Care to guess how the manager will describe your encounter?

“He was a typical chip-head. He never listened to me. He just kept on telling me that what I want won’t work, without once explaining why and without explaining how it is that three of our competitors are doing exactly what I asked him for.”

Is that the way it happened? From your perspective no, from his perspective yes. It’s binary, and his audience is listening to him, not you.

How should you handle a request like this? Be non-binary by responding with a “forcing question” – one that directs the conversation into the right channels. Say, “NT is just an operating system. Let’s hold off on that discussion until we’ve talked through what it is you’re trying to solve. Now what is it you’re trying to accomplish?”

Wham! You’ve forced the conversation into a productive direction. After a while you understand what your new-found friend is trying to accomplish, and then he says, “Now when can I have my NT server?”

Don’t say, “I really think you’d be better off using Linux,” (or Solaris, or AS/400, or Network Computers, or what-have-you). Use another forcing question: “I’m curious – where did you hear that an NT server is the best solution?”

He’ll tell you. He may have heard it from his peer in another company or the sales representative from the application vendor—or it may be that NT is the only operating system he’s ever heard of.

After he tells you, chances are he’ll ask, “Why … is NT the wrong choice?” He’ll ask because the intelligent conversation he just had with you (in which you remained almost entirely silent) has established you as an expert.

If he doesn’t ask, you have more forcing questions to ask, like, “Have you thought through the integration issues you’ll face by going down this path?” He’ll almost certainly either explain his thinking about each issue you raise or ask you to explain it to him.

Don’t take this to extremes. Three issues are enough. Your goal is to move from his proposing a solution to mutual problem-solving.

Let’s imagine all is for naught and this particular manager turns out to be an obstinate fool, insistent on this direction despite your best efforts to bring him to his senses. Do you say no now?

Nope. Your job isn’t to say no. It’s to help this guy succeed. Tell him that. Then use a different negotiating tactic called “absent authority.” Say, “Unfortunately, NT is outside our standards, which means I don’t have the authority to make this decision – let me describe the process you need to go through.”

Make sure the process for approving all bad ideas belongs to the IS Steering Committee, which is chaired by the CIO, populated by the company’s key executives, and responsible for approving resource allocations for IS projects.

Let someone else say no, not you. And if the IS Steering Committee approves this request and is willing to fund it … that’s its privilege.

Yours is to go make it work.

Fistfights sure have changed.

John Wayne only needed one good punch to win a fight, and he was able to take some time to set it up. Thirty years later we have Jackie Chan, who delivers as many as 10 punches (and other blows) per second.

Let’s do some metrics.

The Duke was the clear winner in productivity, achieving a UO-per-punch rate (Unconscious Opponent) of between 1:1 and 1:3 as compared with Chan’s rate of between 1:12 and 1:100. He wins when it comes to efficiency, too – our contender from Hong Kong burns far more calories per UO.

Effectiveness, though, is different from productivity and efficiency. In hand-to-hand combat with a dozen simultaneous opponents, the Duke would have been overwhelmed, while Jackie Chan regularly emerges victorious from duodecimal combatant situations.

Business competition in the Duke’s day looked a lot like his fistfights – fairly slow, deliberate, each move carefully calculated. Today it’s more like Jackie Chan – speed rules.

We all know this. We’ve heard about the acceleration of business cycles until we’re tired of hearing about it. “Internet time” is yesterday’s cliche.

Two other trends — molecularization and service/cost convergence — aren’t yet cliches. Molecularization is the growing need to customize marketing, service, and product delivery — in other words, every interaction — to each individual customer. Service/cost convergence means that customers no longer accept the classic trade-off between quality of service and product price. Combine these two trends with time compression, and every measure of effectiveness changes.

In the old days, companies worried most about measures such as capital utilization ratios, manufacturing output, and defect rates. Their goal was to produce the most products with the fewest defects using the least equipment.

What matters more and more are customer-focused measures such as product utilization (number of products owned per customer), customer loyalty (the likelihood of a customer patronizing your company first), and customer affinity (a customer’s emotional attachment to your company). Other measures, such as shipping service levels — the delay between receiving an order and shipping the product – are equally important.

While the Internet is the most visible driver of this dramatic change in business, the reality is that call centers are still more important in improving these new measures of success.

For companies to succeed in this new world, information technology must permeate every process and activity. It’s possible to “pop” customer information, such as recent purchases and the 10 most recent customer interactions, on the screen of any employee interacting with a customer automatically at the start of any call — inbound or outbound. It’s possible, and your chief marketing officer knows it. When are you going to deliver the capability?

Unfortunately, many IS departments are still mired in the old business model, failing to understand that “late” is just as bad as “buggy” … perhaps worse … when judging the quality of software.

Why? Although some bugs are fatal, others are merely inconvenient. The business will still run while you track them down and fix them. Software you haven’t delivered yet is completely inferior.

It doesn’t run at all.

Correction:

Never confuse a spreadsheet with reality.

As evidence: A few weeks ago I discussed storage options for the 6 billion people now on earth. If you put us end-to-end, we’d only go around the world about 260 times, not the 2,600 times I claimed. I lost a decimal point due to bad parentheses. And a swimming pool big enough to hold all of humanity would have to be about 15 miles per side, not the quarter-mile I stated. One-quarter of a mile is the edge-length of a cube big enough to hold a “Homo sapiens puree.”

Thanks to all who wrote. And no, there’s no ironic tension between these mistakes and my suggestion that “fast” is more important than “bug-free.” These mistakes weren’t the result of haste — I made them quite slowly, in fact. I was just having a bad math day.