While watching Season 1 of Battlestar Galactica on DVD (it’s been a few years since I first saw it), I came across a positively idiotic menu design. You see, there’s a menu where you can turn on the audio commentary. That part works fine. But once you turn it on, you can’t turn it off. Trust me, I tried: I selected the menu option for the commentary again, I pressed the stop button on the remote, and I even ejected the DVD. No matter what I did, the audio commentary kept playing.

Eventually, I figured out a solution: use the remote control to cycle through the audio tracks until the regular audio comes back. Not exactly an intuitive method, but it got the job done. Looking back at the original design, the DVD authoring people basically provided an On button, but no Off button. Though it seems so obvious, the designers apparently never considered that someone might actually want to turn off the audio commentary.

We can generalize this into a useful rule of thumb. Simply take a look at each of the options that you let your customers turn on, and make sure you give them a way to turn those same options off as well.

I installed some backup software over the weekend, and it reminded me of the reasons why traditional computer software sucks:

– Updates are a pain in the ass and the automatic ones never work right.

– The user interfaces are typically very crude and confusing when compared to web-based software.

– The pricing models don’t match the value delivered. I would much rather pay $30/month for a product that I use all the time and get support and upgrades with that, rather than paying $400 upfront and $200 every few years for the upgrades. Why? Because the web pricing model is less work for me (no upgrades to keep track of or waste my time purchasing and installing), and I pay more directly for the ongoing value the software provides me.

What does this mean for desktop software makers? Web-based software has raised the bar in many areas, and I believe traditional software needs to improve quite a bit in order to stay competitive.

If you’ve ever used a Polycom IP phone, you may be aware of a somewhat strange behavior. When you pick up the handset, you hear a dial tone. The problem is, you hear the same dial tone nearly all the time — even if the network connection is down and the phone can’t make calls. (The only exceptions seem to be if the physical network cable is unplugged, or you try to reboot the phone without a working network connection.) In this fashion, Polycom’s implementation of the dial tone sound violates the user’s expectations. A dial tone is supposed to give you the all clear to make calls, but now the sound plays even when it’s impossible to place a call.

More generally, the situation can be described as follows. The designer of the new version of a product (or an entirely new product replacing an old one) makes the smart decision to emulate the familiar behaviors of the old product. They employ visual cues, messaging, and sounds that help users become comfortable with the newfangled solution. But when some of these legacy characteristics don’t really apply anymore, or the old attributes are copied poorly, you get a mismatch.

When integrating these types of elements into a new product, the best approach is to think carefully about what each one means to the user and what information they expect it to convey. In the case of a dial tone, the solution would be to only play the noise when the IP phone has a working network connection and is communicating properly with the PBX server, thus restoring the value of the dial tone as a status indicator. By thinking this through, designers can make new products that are familiar and intuitive to users, even when the underlying technology is vastly different than what came before.

Updated on August 23: I added more details about the dial tone behavior on the Polycom IP phone, since there are some cases where it actually does stop playing the sound.

At most supermarkets, you can choose which type of bag you want: paper, plastic, or both. You can even bring your own bag if you’re so inclined. But these choices are entirely absent from online shopping, and the packing materials tend to be pretty random as well.

For example, my most recent Amazon order came in two boxes. The first was filled with brown paper, while the second contained those inflatable air pouches. I don’t know which of these is more environmentally friendly, but I’m sure somebody has thought it through. Assuming the costs aren’t much different, it would be great if Amazon let you select your preferred packing materials when you place an order.

With all the options and customizations available to consumers today, I’m surprised that online stores don’t let you choose your packaging method. You can already get things gift wrapped on a per-item basis, so I’m guessing the infrastructure is there to support individualized packing preferences. Maybe the benefits to consumers are dubious, aside from the obvious benefit of choosing the most environmentally friendly packaging. But I’m curious how customers would respond if someone like Amazon started letting you choose how your orders are packed.

If your so-called “web application” only runs in Internet Explorer, it’s not a web application. Rather, it’s an “Internet Explorer application”. I was reminded of this simple fact while looking for a new software package the other day. One of the products on my short list was web-based and looked pretty solid. But when I went to the demo page, there it was: “Sorry, you are using an unsupported browser. You must use Internet Explorer to access our site.” Needless to say, I crossed that one right off the list.

I don’t know which planet that company was from, but a web application needs to work with the popular browsers that people use. Today, that means both Internet Explorer and Firefox, and maybe Safari if you’re targeting Mac users. By only supporting a single browser (even if it’s the most popular one), you’re limiting your market penetration. And calling the resulting product a “web application” only makes the situation worse.

Maybe it’s the type of sites that I’ve been looking at lately, but the “Contact” link seems to be disappearing. Meanwhile, the contact page is getting buried several levels deep in weird places like Company or About Us. Or, they’ve moved it to a hard-to-find spot at the bottom of the page.

Perhaps I’m old-fashioned, but I believe visitors should never be more than one click away from getting in touch with you. Sure, you can decide how best to handle those inquiries, or even go the email-only route. Whatever approach you take, featuring your Contact link prominently on each page provides a subtle reminder that you stand behind your products — and shows customers that help is there when they need it.

I’ve noticed a lot of variance in how retailers report the sizes of their items. Usually, the dimensions shown will reflect the product itself, but sometimes they’re actually describing the shipping box or a carton that contains more than one unit of the product. Worse, for items that have an inside measurement or carrying capacity (such as luggage or coolers), the data is even harder to figure out.

I propose a simple solution to this: just label each set of dimensions so people can understand them. For virtually every product, this means there will be a line labeled “Product dimensions”. If the item has an inside compartment, you’ll also have “Interior dimensions”. And finally, if the size of the shipping box is available, call it “Shipping carton dimensions”. For cartons that contain more than one item, specify that as well.

Sure, this all seems totally obvious. But I rarely see dimensions stated clearly, even on major sites like Amazon. By adopting this simple approach, retailers would significantly reduce the number of returns due to products that were much bigger or smaller than expected. At the same time, they would sell more products by eliminating the need for customers to leave their site and hunt all over the web to find the real dimensions of what they’re trying to buy.

Yesterday, I needed to make a purchase from a site that I’ve been doing business with for years. I added the products to my cart, proceeded through checkout, and submitted the order. Then things went south. The site generated a cryptic error code, and asked me to call customer service for help.

I dutifully followed the instructions, and waited on hold for about ten minutes. Once the customer service rep picked up, he immediately found the issue: the credit card was declined. It turns out they had an old card on file, so I replaced the card in my account with the new one, and the transaction went through fine.

So what’s the problem? Well, I had to spend ten minutes on the phone to deciper an error code that clearly meant the credit card transaction failed. The site should have just said “We had a problem processing your card. Please check the card number and expiration date and try again”. This would save time and money for them, and make things much easier on customers.

The moral here goes something like this: if you have information on why a problem occurred, then tell the user about it. Don’t make them call you to figure out what the code means. Even if you have to give three or five or ten possible causes and solutions for a given error code, you’re at least providing customers with a way to resolve the issue quickly — on their own terms.

Even when my company isn’t advertising any open positions, we get a steady flow of unsolicited resumes. One type of resume is particularly strange, though. The subject line reads “Resume of John Smith is attached”, with “John Smith” being replaced by the applicant’s name. The message doesn’t mention any specific position, and the resume itself is similarly untargeted to our company and industry.

I’m guessing that these rather clueless individuals hired a placement firm to find them jobs. The placement firm then uses the same template for all their outbound communications. I once received three messages based on the template in the span of a few hours, so this scenario seems likely.

If you’re working with a placement company that uses these tactics, be aware that you’re part of the problem. Their senseless spamming makes you look guilty by association, and probably results in very few actual job interviews. So, when looking for someone to help you find a new job, make sure you understand their tactics before you sign up. After all, if a company is gullible enough to fall for these spammy tactics, do you really think they’re going to provide a good work environment?

I’m not very well-versed in the area of webinars and online meetings, although it seems like a lot of people are doing them lately. The increasing cost and hassle of air travel is probably the main reason for this trend. But since these virtual meetings don’t require any travel, do potential attendees really need much advance notice?

The approach that meeting planners have carried over from the days of physical meetings is to notify people months in advance, and keep hammering them with reminders until the day of the event. But I think a simplified approach can be even more effective: tell people about the meeting maybe a week in advance, and send a reminder again the day before it starts.

I recently viewed a webcast that used the just-in-time approach, and the attendance was outstanding. It’s counterintuitive, but maybe today’s short attention spans mean that less notice actually increases the chance of people showing up.