Too many companies have forgotten that automation is intended to take care of the ordinary class of problem-and that humans are a necessary ingredient for uncommon scenarios. The result is poor software usability, particularly for the developers, techies and admins who actually do know how to operate a computer.

Let me give you a fer-instance.

My personal (non-work) mail server is set up with rather strenous levels of spam-fighting, so I sometimes get complaints from people who have misbehaving mail servers, and the occasional bounce from poorly run e-newsletters. It usually isn't a problem, once I reach the right individual, since it can be resolved simply by the user (who ought to be taught to send mail properly in the first place) or by the admins (who may not know they're breaking the rules; ignorance, fortunately, is curable).

But sometimes, there isn't a human available to correspond with. I received the New York Times e-newsletter for years, but it unaccountably stopped arriving. We've checked the mail server logs; the New York Times newsletter hasn't even attempted to deliver for several months. The company's site has no way to tell them so. One of my credit card companies insists that my e-mail ID is invalid; in actual fact, Capital One's SPF record is broken but the site has no way to tell the company that they're the bozos.

Similarly, [US Online DVD rental company] Netflix has decided that my e-mail ID is invalid, and regularly displays a message direly instructing me to fix the situation immediately. (Not that it keeps movies from arriving at my door. That still works.) Netflix's Help text is written for ordinary users, for perfectly good reasons:

If you get this message, it means that we are not able to contact you via the email address you have provided us. You may want to check and update your email address. If your email address is correct, you should add Netflix to the 'safe senders list' in your email or spam blocking system. The addresses you should add are info@netflix.com and discship@netflix.com.

And I'm cool with that. For most end-users, that's the right message, even if it places the blame on the user. My issue is that there's no place to click when your response is, "Tried that. Didn't work." (And, incidentally, Netflix.com has not attempted to send any messages whatsoever for the last few months, although I did re-enter my e-mail ID just-in-case.) Netflix kindly gives me a telephone number for me to call, but frankly, I rarely pick up the phone these days; and that's not the point, anyhow.

What I want a site to do is offer the basic user information, such as the text that Netflix offered above. But there should be a link for "I tried that; now what?" that bumps me to the next level of support. When an error message says, "Contact your administrator," I should also see a link for those who say, "I'm the administrator, and I sure can't figure out what the problem is." That is: I want a way for a technical person to contact a technical person. Too few sites have no way to do so.

Since those companies are the ones send out bulk e-mail (solicited, in this case) they are also the ones most likely to be added to a black list somewhere. The fact that their errors and fixes tend to blame the user rather than to say, "Here's where to go to let us know we may have a problem" is a serious weakness.

This isn't the fault of the software development team, who is just following the site spec. It is, however, a problem with the way that too many sites are designed. And that's just plain wrong.