Easily Startled

I have never met anyone who claims to design interaction flows for telephone menus. This could be because people who do are too ashamed of their job to ever mention it in public, or it could be because the menu trees are all designed by small monkeys in a hats.

If said monkeys are reading this, I would like to suggest the following rules when designing your flow:

1. Don’t ask me for my phone number. Or if you have to, make sure it is done as a last resort. If you can get my phone number from my caller ID, or my account, please do. Typing in your own phone number and having it repeated back to you in a slow, monotone is painful.
2. I KNOW your menu has changed, I am LISTENING CLOSELY, please stop telling me that. I don’t just punch random keys, I promise. If you are never going to update your message, don’t make it something that talks about something being ‘new’ or ‘changed’.
3. On a related note, I can only imagine you never expected anyone to actually dial your number, since any call volume seems to be highly unusual to you. How about you skip the surprise and just tell me how many days I can expect to wait?
4. Don’t make me press a number to end the call. I can end it quite well by just hanging up. Menu choices like “If you are done, press 1″ are silly. There is one big red button I hit when I am done.
5. Look at the default case, can you remove any key presses? Maybe make the default action just to continue on ahead? If so, remove the extra steps.
6. Don’t start a menu tree choice out at 2. Or 3. The first choice is 1.
7. If 0 is cancel, let it always be cancel, don’t move it to 2 sometimes and 3 others.
8. Model towards a web browser. Give us a back button (like 0) so we can walk back through the tree if we feel we have made a mistake.
9. Use your own product. A lot. Make sure you can handle it, because if you can’t then all the other monkeys of the world aren’t going to like it any better.

I actually think there would be some interesting challenges in phone menu design, and it has the benefit of being a pure information architecture problem.

There is no sport in pointing out flaws in Microsoft Office products. In fact, in their defense, they have the difficult task of continuously adding new features to a set that was complete 10 years ago. When the pressure is on to shoehorn in 3-5 new back-of-the-box of features to a yearly release, complexity is a way of life. IT staff don’t buy software based on “20 useless features removed!”

My latest UI annoyance came after PowerPoint repeatedly requested that I center my bulleted list. There was clearly a default setting for the slide which was to center any new text element. This came into direct conflict with my desire to add a bulleted list. This is unfortunate. I can think of no reason to EVER center align a bulleted list. That pretty much defeats the whole purpose of a bulleted list.

Which brings me to my UI rant of the day. Be context aware, please. There are few times I get more frustrated with an interface than when it becomes clear that the application’s left hand has no idea what the right one is doing. It can be challenging, for sure, but this is what interaction designers are tasked to handle. Build that flow chart, map out the states, try to anticipate what the user is doing, and act like you are there to help. Ironically, it doesn’t take much to look smart.

I thought we had an agreement here, guys. Somewhere in the Hippocratic Oath for UI designers we all agreed to only use red for error messages. Now I feel that some of my brothers and sisters have gotten off message.

I saw the above example today, letting me know: “OMG! YOUR PAYMENT! IT WAS RECEIVED!” Which is fine, but it took me a second to realize that, in fact, everything was OK. Would not a green also have been attention grabbing and ignited less of a fight-or-flight instinct in the user?

Whole Foods has installed a nifty new credit card reader that, after processing your payment, flashes a big, red “ACCEPTED!” I still do a double take. What did it just tell me? Is everything OK? “Your credit card was not REJECTED! Don’t worry, YOUR BANK ACCOUNT has not just been reduced to ZERO FUNDS!”

Let us go back to keeping red for non-normal states. Trust me, there are other colors that will stand out almost as well, and, at least in the US, they won’t imply emergency.

Thankfully touchscreens are making this particular ATM problem a thing of the past, but I do feel that this is a prime example of why you should test your system in the environment in which it will be used.  In the lab, I have no doubt that these prompts and buttons line up great.  However, add in a glare and a slight angle, and suddenly the choice isn’t so clear.

Even ‘free-range’ lab testing is no substitute for real-world observation.  It is right to ask, as Schrödinger once did, if it is even possible to observe a ‘natural’ interaction?  Or does our mere observation always taint the data?  Probably, but it is worth the effort to try.

I feel the hotel is sending me mixed messages with this one.  On the one hand, the red strike-through cigarette is says: “Don’t Smoke.”  While the ashtray which sits upon it says: “Go right ahead, smoke in bed, burn holes in the sheets, just don’t get any ash on the floor.”  I guess they are just being realistic.

A friend of mine sent this along as another stellar example of instructive content.  The best part of this photo, to me, is an image of a increasingly frustrated parking lot owner, who keeps adding more signs to the machine.  Clearly since the last 3 he added had no effect, the solution must be to add one more.  The three “step 3″s is amusing as well.

You probably need enlarge this one to get the full effect. A co-worker of mine likes to say: “somethings just don’t need a label.” I argue that this might be a good example of that.

I found this bathroom gender identification sign in a hotel a little bit odd. Is it just his pose seems exaggeratedly “manly”? Or does the angle of the legs look a bit too similar to the “skirt” of the female sign? Maybe I was just jet-lagged, but it made me question the choice I was about to make.

By Paco Underhill

Why we Buy provides a detailed look at how people shop for everything from jeans to jumper cables. Underhill’s Gorillas-in-the-Mist approach to shopping research proves fruitful as he examines every intricate detail about the allusive beast known as the “shopper.” While not everything directly translates to interface design, his dedication to total understanding of the user’s habits is admirable.

Oddly, while reading the book, it is hard not to picture Underhill as a tweed-clad anthropologist frozen since the 1950s. Something about his choice of phrases or his fascination with gender differences gives the book a slightly dated feel. None of it is done in an offensive way, but more in that matter-of-fact tone that suggests nothing could be more widely known than that women don’t feel comfortable shopping in hardware stores. It sports a Leave-it-to-Beaver charm.

Despite this quirk, I recommend this book to anyone interested in commerce, ‘e’ or otherwise. It is also fun to apply the critical eye gained from the book to your daily excursions to the grocery store. And the book is not without practical lesson for us web types. I think that ideas such as the “butt brush effect” might actually have a digital equivalent. Underhill describes the effect as the drop in sales that occurs when aisles are placed too close together. Shoppers who are “brushed” by other shoppers trying to get by, are much more likely to feel uncomfortable and leave without a purchase. While the cause of the discomfort is different online, it is easy to see how even small frustrations or awkward moments could prevent a web purchase. What can we do to make sure people feel comfortable and safe all the time while on our site?

Underhill discusses transition zones and how their presence at the entrance of a store is important to allow shoppers to make the adjustment from ‘parking lot’ to ’store interior.’ He suggests that leaving that zone empty might actually be the best use of space one can make. I wonder if homepages should functions as such transition zones. Do we need a landing pad for the highway surfers skidding from one site to the next? But if so, how one deals with visitors hitting item pages directly? Unlike a physical store, visitors can skip the entrance and jump right to the shelf they want. Do we still need to find a way of easing them into the shop?

Ironically, the author himself doesn’t put much stock in online shopping. He feels the sensory experience is significantly lacking in the virtual store. Of course he writes all this nearly 9 years ago, so he lacks the benefit of watching sites like Amazon and Zappos really take off. It is too bad, as I would have liked to have read his take how these offline observations might pertain to the online world. As it is, I can only hope to make my own.