Bad design: Deliberately restrictive sales

This one is old as the hills, but tech companies are still trotting out the “enticing sale with decent discounts that excludes almost everything you’d actually want to buy” offer. This particular offer also has a bonus loot box component where the promotional email gives you a random code for a discount between 10-25%.

In this case, the offer came from Logitech. My alleged discount was 25%.

I clicked the enticing CLICK TO REVEAL button and was given the 25% off code. I’ve been thinking of acquiring a full keyboard with keypad again for times when I might want the keypad and the Logitech Craft gets good reviews–but is also ludicrously expensive, selling for around $200 Canadian most of the time. This discount would bring it down to a more palatable $150.

I then read the not-so-fine print at the bottom of the email that lists the items excluded from this offer (remember, the discount starts at a not-exactly-gigantic 10% off). There are not 33 items on the list (which would already be a lot), but a combination of 33 individual products and entire product lines:

You can buy anything that is old and cheap, however.

The Craft keyboard is among the impressive list of exclusions. Almost anything new or on the pricier side has been left off the sale. Why? Because Logitech wants you to pay full price for those. Perfectly understandable. For profit companies like profits.

But this promotional offer–even if you overlook the skeevy loot box “What did I get?” aspect still stinks. The unspoken hope here is that the potential buyer will not read over the list of exclusions and if they try to buy something that isn’t part of the promotion, well, they’re already on the Loigtech site, so maybe they’ll end up buying something else. Or shop around and get ideas, even if they don’t buy something right then. This is questionable marketing at best, and dishonest at worst. It trains customers to not trust you when you offer something. If I got another offer from Logitech, I’d immediately ask myself, “But what’s the catch?”

But that won’t happen, because I’ve already rewarded Logitech by unsubscribing to their offers. Well done, marketing geniuses.

(The unsubscribe option is simple and didn’t even ask for a reason, which is too bad, because I wanted to tell them!)

Lousy keyboards of yore

The Wall Street Journal published a column today by Joanna Stern in which she reports that Apple’s butterfly keyboard used on its MacBook, MacBook Air and MacBook Pro laptops is still having issues three generations in. This prompted Apple–currently facing a pair of class action lawsuits over the design–to offer an apology of sorts:

“We are aware that a small number of users are having issues with their third-generation butterfly keyboard and for that we are sorry. The vast majority of Mac notebook customers are having a positive experience with the new keyboard.”

Apple didn’t say they were sorry for first or second generation butterfly keyboard owners, likely because every one of those keyboards is guaranteed a free keyboard replacement up to four years after purchase.

Apple has effectively admitted there are issues with all three generations of the butterfly keyboard. I have gone from hating the feel of the keyboard (mine is the dreaded first generation) to tolerating it. I’d prefer to have more travel on the keys and have them be quieter/less clicky, but could otherwise adapt to them. The third generation, with its silicone membrane is apparently a little less noisy, but I’ve yet to test it out in a quiet-enough environment to notice a difference. Also, the membrane apparently contributes to heat build-up, creating a new avenue for issues to arise.

All said, what John Gruber calls “the worst products in Apple history” are perhaps hopelessly flawed. I mean, if issues are still coming up after multiple fixes, maybe it’s time to move on to another design entirely?

The MacBook is overdue for a refresh. If Apple doesn’t kill it, the next version of it may show if Apple is staying all-in on what appears to be a fundamentally broken design, or gives up and goes for something else, like adapting the low-profile scissor switch design used in their external keyboards for their next generation of laptops.

I’m leaning toward the latter at this point, mainly because of today’s apology. It feels like the beginning of the groundwork to kill the butterfly design and bring in something butterfly-like, but with none of the fragility.

And while reading about this today, I came across PCWorld’s The 10 Worst PC Keyboards of All Time. The butterfly keyboard isn’t on the list, as it dates all the way back to 2007. Still, it’s a fun–and horrifying–read. It’s kind of amazing how many computer keyboards didn’t have a backspace key.

Bad design: Relying on images for words

A featured deal on the Kobo eBook site looks like this (this is not clickable and the deal may have ended ten years ago by the time you see this, so apologies if this was your most-anticipated book ever and you can’t buy it at the price shown):

Seems fine, right? Now imagine if for some reason the site wasn’t rendering the image of the book properly. Here’s the same ad, with the book cover obscured:

No big deal, right? But let’s say you don’t have time to click the Learn more link when you see it, so you make a note to check later. But later the ad is no longer in rotation. Still no big deal, because you made a note of the title of the book–oh wait, you didn’t, because if the image doesn’t display, the text has no mention of the book title anywhere. Whoops. And this is applied inconsistently as well, as here’s another featured ad from the same page and it does mention the title in the last paragraph:

This isn’t world-ending bad design, of course, it’s really more of a nitpick. Still, an ad for a book should perhaps mention the book, no?

Bad Design: SkyTrain stairs and elevators

There are various issues with the location and design of stairs and escalators across the entire SkyTrain system and I have posted about issues specific to the Canada Line, but for this I am going to focus on the station closest to where I live, as it is a good example of bad design.

First, let me note that the Millennium Line, opened in 2000, is overall an improvement to the Expo Line and its stations. The Millennium Line stations are spacious, completely covered for inclement weather (important for an area that gets a lot of rain), glass-enclosed elevators for better security, and each station has its own unique look, getting away from the cooke-cutter design of the Expo Line.

One area in which they went cheap was escalators. A lot of stations follow the one staircase/one escalator rule, where there is an up escalator to get people up to the platform, and a staircase to get them down. As cost-cutting measures go, it’s not the worst, but they are moving away from it now, because crowded stations and stairs are inefficient for getting people in and out quickly (see Lougheed Town Centre station, which added a down escalator at the north end, around the same time its third platform opened for the Evergreen extension).

At Sapperton station, both sides feature the one staircase/one escalator design. The problem here is that designers did not anticipate how people act. A lot of people will—quite logically, you could argue—take the path of least resistance. In this case, when someone exist a train, they will veer toward the closest route that will get them off the platform and out. At Sapperton this is the staircase, as it is closer to the platform than the up escalator. The up escalator requires the person to cross over the platform.

Now, you could argue toe the logic of this design is that having the staircase closer to the platform is more important, because it makes it easier for people to board a train. And that’s true. However, by putting the staircase closer, you inevitably increase the number of people using it and ignoring the escalator, which is farther away.

The end result is you get people both going up and down the stairs. This impedes people trying to get to the train, the very thing the designers were presumably trying to avoid. It also creates cross-traffic on the platform itself, as people exiting a train and using the up escalator must cross in front of the stairs.

During busy times, it’s a bit of a mess.

Now, if the stairs and elevator were reversed, you’d encourage people to take the escalator to leave the platform, plus you’d keep them from getting directly in the path of people coming down to board. People coming down would have to move slightly farther to get to the staircase at the top, but this would almost always be more efficient than putting them into the direct path of people leaving the platform.

Unfortunately, this design will never get fixed in existing stations and really, it’s too expensive to be worth the improvement in traffic flow, as nice as it would be. More happily, as mentioned earlier, it appears Translink is largely scrapping the use of stairs. An example is the remodelled Metrotown station, which previously had a single staircase and escalator at its east end. It was a huge bottleneck. The remodel opened up the west end of the platform and now there are four escalators (two up, two down) at both ends, vastly improving the efficiency in getting people in and out of this often crowded station.

So this is a case of bad design, but bad design recognized. Thumbs up, I say.

Bad design: The Canada Line

The Canada Line is the ALRT that runs from downtown Vancouver to Richmond, splitting off to either the airport or central/”downtown” Richmond. It opened for service in 2009 and one of its intended functions was to help move people from the airport to Olympic venues during the 2010 Winter Olympics.

Obviously the construction of a subway is always going to be disruptive to the neighborhoods it goes through, but the Canada Line’s first bad design was the decision to use cut and cover tunnels for most of its route. Cambie Street is the main route the line follows and it literally follows it–being cut and cover, they dug up Cambie Street, keeping two of its six lanes open during construction–and laid down the subway route along the road, following the same route, including a prominent “bulge” around Queen Elizabeth park. Businesses sued over the construction, some went under or moved and in the end it was declared such a mess that the entire Millennium Line extension from VCC-Clark to Arbutus is going to be a bore tunnel.

But the construction was only the first tangible aspect of the bad design. Once open, the line revealed its other flaws to the public. Namely, it was built on the cheap and with no thought put into expansion.

  • The system only runs two-car trains, severely constraining capacity (it’s already maxed out at rush hour)
  • Stations are very small and often have multiple bottlenecks, leading to serious crowding issues
  • Fare gates are often too close to the platform, again because of the tiny size of the stations
  • The small stations means that even if they expanded trains to four cars–which they could do anytime–there would be no room for the trains at any of the current stations. They would literally have to dig out new space everywhere except for the above-ground stations in Richmond. Compare this to the Expo Line opened in 1986, where all platforms can accommodate six-car trains, even though they only ran four-car trains for many years.
  • Adding insult to injury, the Canada Line even started removing seats from some cars to allow more people to squeeze in, a terrible solution to overcrowding that will only drive (ho ho) people to seek alternatives. They fortunately stopped this after gutting a couple of cars, but the fact that they did it at all shows how inept the leadership is.
  • The terminus Brighouse station in Richmond is specifically not designed to accommodate any expansion of the system. When the train leaves the previous station, Lansdowne, it switches over to a single track. The Brighouse station is built specifically around that single track, meaning any expansion would require a massive and expensive retrofit. Compare this to the Millennium Line, which anticipated a future expansion to the Tri-Cities called the Evergreen Line, leading them to build the base infrastructure for a third platform at the Lougheed station, where they expected the Evergreen Line to hook up to the rest of the system. When funding finally got approved for Evergreen–more than a decade later–crews were able to build on and around the existing third platform, saving time and money.
  • The trains are seriously noisy when going around corners, the wheels shrieking like crazy. This does not happen to the same degree on the underground sections of other lines.
  • Even the choice to go with a Korean manufacturer was short-sighted, as it meant the system could not integrate with the two existing SkyTrain lines, perhaps saving money in the short term, but hobbling any overall integration in the long term.

Much of this can be laid at the feet of the former BC Liberal government, who wanted the line built cheap and fast. They got both of those, but what they didn’t get was the unspoken third part–good.

Fixing the Canada Line would be difficult and costly, but it will likely have to be done eventually, simply because it’s already reached its capacity. There’s no happy ending here.

Bad design: The easily-circumvented no-biking-for-you gate

Usually at lunch I walk the loop around the Langara Golf Course. It’s about 2.7 km total and is pleasant enough if the weather is decent. The path does not allow bicycles and there are a few signs to that effect at regular intervals.

Recently the point was made even clearer through the erecting of a pair of gates that are tricky for pedestrians to navigate, but downright hostile to cyclists. These are meant to discourage and turn back cyclists and in theory they should do just that.

Here’s one of the gates.

To bike: You shall not pass. Bike: Well, actually…

You can clearly see some people are winding their way through the gate. You can also clearly see enough people are walking around it to create two distinct side paths, one on each side. I am one of those people.

I have yet to see a cyclist navigate around this gate, but I’m reasonably confident that in time I will.

The other gate is up against a fence, so only one side can be passed on the outside, and it’s narrower, but people are working on it.

This is bad design.

These are probably prefab pieces or otherwise made to a standard size. But if the actual location they’re being used in makes them trivially easy to avoid (and it is trivial–in the one pictured above a jogger can deke around the gate without slowing down), there’s not much point, really. You’re just adding an inconvenience for all the people who walk the trail.

My preferred solution in this case would be to remove the gates. Unless there is a serious problem with constant crazed cyclists, there’s little value to be had in placing these obnoxious, easy-to-avoid barriers.

Bad design: Paving over the (wrong part of the) environment

(Technically this is not paving, but pouring concrete.)

Here’s something you sometimes see in cities or any urban area that has decent foot traffic. For the most part people walk along sidewalks and stick to them because the alternatives are unattractive for varying reasons–trespassing in someone’s yard and getting eaten by a dog, walking into the street and being run over by a bus, and so on–but sometimes there are routes off a sidewalk people will take, when that route is more direct and without risk of dogs, buses or other obstacles.

This is the sidewalk in front of the Burnaby Public Library on Central Boulevard.

Walk which way?

There are a pair of roundabouts here that bisect the sidewalk. Here’s an image from Google Maps that shows both of them (the one in the photo I took above is on the right):

A decision was made to have the sidewalk in front of the library not line up directly along the street, as would normally be the case. As you can see especially from the photo, people who come from either intersection, where the sidewalk does align next to the street, continue to walk in a straight line, ignoring the sidewalk directly in front of the library altogether.

The unnatural foot path (so to speak) becomes quite muddy and slick in winter, but it’s perfectly fine for most of the year, provided you’re not especially susceptible to tripping on small, exposed tree roots.

Anyone who understands human behavior should have anticipated the development of this foot path when looking at the plans for this block. Sidewalks that deliberately steer you away from getting from Point A to Point B are bad design. People will always take the shortest route if it’s easier and safe to do so.

The fix here would be to add a stretch of sidewalk where the existing foot path lies. The tree roots complicate things, so that may never happen, but the roots will eventually get ground down by the traffic on them, anyway.

Sometimes this bad design, while not anticipated, is still fixed later. Such is the case in Thornton Park in Vancouver. When the park was rehabilitated as part of the redesign of the nearby Main Street SkyTrain station, several pedestrian-made paths were converted into permanent sidewalks. It’s always nice to see recognition of real-life usage and adapting the environment to it where it makes sense.

Bad design: The iMessage Fireball

In the interest of keeping to a complaint-free lifestyle, I’ll emphasize again for Bad Design I am doing a couple of things:

  • pointing out the bad design as a way to highlight how something should not be done, even if it seems logical or a popular way to go, in the hope that it encourages others not to repeat what I feel are mistakes in design
  • offering a solution or alternative design that addresses the flaws

And I only pick a lot on Apple because I own and use a lot of Apple products (which I will address in another post) and because as the world’s largest, richest company, they have the power to influence a lot of others (see Samsung and its weird and lawsuit-attracting tendency to follow Apple’s designs very closely).

And with that, I present:

The iMessage fireball

For those unfamiliar with Apple’s message app, it works like most text messaging apps and allows people to send and receive messages across Apple devices, including the iPhone, iPad, Watch and Mac. If you send a message to a non-Apple device, it shows in a green bubble as a regular text message. If you send a message to an Apple device, the bubble turns blue and it becomes an iMessage, sent through Apple’s servers.

With iOS 10 Apple revamped the Message app, expanding what you can send.

When you are in the Message app, tapping on the App Store icon presents a small black window that you can doodle and do other things in (it defaults to this, though tapping the icons to the left or right of the heart will allow you to use stickers from other apps, search for images and more):

By tapping on the horizontal gray line above the window it expands to give you more room and also exposes a small information icon in the lower-right corner that, if tapped, presents an explanation for the various actions you can perform:

The Heartbeat option only works if you are wearing an Apple Watch, as it includes a heart rate monitor.

The first option is Sketch and it seems pretty straightforward. Draw with one finger.

The next option is Tap. What is a tap? There is no explanation.

The fourth option, Kiss, puts a pair of lips in the window.

Let’s go back to the third option, Fireball. This puts an orange fireball-like blob in the window and as long as you keep pressing you can move it around. As soon as you release, the Fireball message sends.

Bad design #1:

Some actions send the message instantly, others require you to tap a send button to send it. This is inconsistent and can result in messages being sent prematurely.

Bad design #2:

There is no explanation for what a Tap is. The others are straightforward, but what is a Tap? It’s a ring that dissolves. If you do a bunch of taps in succession you can send multiple rings–er, taps–but if you pause too long the message auto-sends. I am unsure as to why anyone would ever want to send a Tap.

Bad design #3:

The Fireball and how it is invoked. Like the Tap, I’m not sure why you would send someone a Fireball. It looks more like a glowing orange ball than an actual fireball and it also auto-sends. The worst part, though, is that to invoke it you press your finger on the screen. You might think this is the same thing you do to make a Sketch, but there is a subtle difference. The difference is so subtle that you may find yourself sending off fireballs when you meant to start a sketch, and you may receive fireballs for the the same reason. In fact, since iOS 10 launched I have only sent two fireballs deliberately. The first was to see what it looked like, the second as a joke. Pressing the screen is a very basic gesture and it shouldn’t be tied to a fairly obscure action that few people would seemingly ever use.


Bad design #1: Require tapping the Send button for all actions before the message is sent. Give options to edit or cancel the message.

Bad design #2: Rename Tap to Rings. Change text to “Tap with one finger to place one or more rings.”

Alternate solution: Remove this option altogether if it is seldom-used.

Bad design #3: Change the action required to invoke the Fireball to something that is not likely to be used accidentally, like tapping with three fingers.

Alternate solution: Remove this option altogether.

My personal feeling is the Tap and Fireball options could be removed. I have no evidence to back this up, but based on anecdotes and my own experience, neither is used much at all and the Fireball is almost exclusively used unintentionally.

Bad (but logical) design: Bike spaces on SkyTrain

The newest Mark III SkyTrain cars feature a few nice improvements:

  • all four cars are joined together through an articulated “accordion” section, meaning you are free to move between all cars on the train. This also means there is more room overall for passengers
  • larger windows provide a better view and the lower frames work better as pseudo arm rests
  • better fittings all around mean the trains are quieter
  • roomier design all around, so there is less of a sardine can feeling, even when the train is full

But in among these improvements is another that doesn’t really work, and it’s not the fault of the designers. It’s more of a people problem.

The first and last car on each train has one of the middle sections of seats removed and in its place a single bar that runs underneath the window. This is a designated bike area. Making trains bike-friendly is definitely a nice move, as more people are commuting by bike.

However, there is a problem with the execution: it doesn’t take into account normal human behavior and the general likelihood of bikes being on the train at any given time. This leads to the following:

  • as the train fills, people move first into the seats
  • they next stand in the areas that are most open (not between seats), such as the doorways
  • conveniently the bike space is wide open, so it often fills up with standees before the rest of the train
  • a cyclist boarding the train at this point will find it impossible to park their bike in the already-occupied space. Even if people wanted to let them, it’s unlikely there is room for the people in the bike space to stand elsewhere; the cyclist typically props their bike in the doorway area, same as they would without a bike space on the train

Given that cyclists are still uncommon on the SkyTrain and that they have no better chance of boarding before anyone else, there is only a small chance they will actually get to use the designated space for their bikes. There’s also no way to keep other people out of the space (nor should there be).

Conclusion: the dedicated bike area is a well-intentioned idea that ultimately doesn’t work. It’s really just a standing room section that would be better serviced by putting the seats back in.

However, there is a better solution that, while still subject to the whims of the crowd on the train, at least doesn’t remove a bunch of seats. Some rail systems have hooks in the ceiling that bikes can be hung on. This works well for a couple of reasons:

  • the hooks aren’t likely to be used for something else, so they will almost always be free for cyclists to use
  • the bikes stand vertically as a result, taking up a lot less space on the train
  • the hook provides a solid anchor for the bike, reducing the chance of it hitting someone or getting away from its rider

I hold out hope that Translink will ultimately switch from the dedicated space to a hook system and am doing my part by suggesting it to them, not just here, but directly via email as well.

Until then, the bike space on the Mark III trains is likely to remain standees-only.

Bad design: This donut

It is a well-established fact that I love donuts. I love all kinds of donuts, too–glazed, cake, jelly, pretty much any donut is good (except maple donuts. Maple and donuts just don’t work together for me. It’s like combining peanut butter and chocolate, but instead of peanut butter you use plaster and instead of chocolate you use motor oil).

However, this donut is wrong.

No, not the decadent but possibly decent Chocolate Cheesecake Donut on the left. I’m talking about the one on the right.

The Angel Cream Donut.

Tip: If your product is using the words “angel cream” you are very likely doing it wrong. The only thing that visually distinguishes this donut from a Boston Cream (mmm, Boston Cream…) is the white icing drizzled across the top. You know, the angel cream. Or maybe the angel cream is inside the donut and it is meant to simulate a pureed form of angel, whipped and blended into a horrifying but richly smooth cream-like substance.

I don’t know. I don’t want to know. The fact that the sign is handwritten suggests this could be a rogue donut named by capricious staff. More likely the official Angel Cream Donut signs haven’t arrived at the store yet because the person who is printing them keeps looking at the sample and going, “Ew!” and never prints anything.

Bad design: Soylent green is people (not physicians?)

The Dragon Naturally Speaking site has a Solutions section at the bottom of the site that seems to suggest that physicians are not people:

Then again, it does separate business and people, which makes sense, as people are not businesses.

The real problem, of course, is using “people” because no matter which version of Dragon Naturally Speaking is chosen, it’s a pretty safe assumption that it will be used by people rather than cats, robots or giant carnivorous plants.

The solution would be to replace “people” with something that more accurately reflects the product:

Speech recognition — for individuals
Speech recognition — for business
Speech recognition — for medical use

Note that I also changed “physicians” to “medical use” since you kind of need to be in medicine to be a physician and this better aligns with “business” being the other non-individual choice. Note also that the link for Speech recognition – for people actually leads to a page offering Dragon Professional Individual so I’m wondering why I even have to suggest this change in the first place.

Finally, note that there is no way to easily see a list of features to differentiate the many flavors of Dragon Naturally Speaking. What makes Home different than Premium, other than the latter costing $100 more? Premium obviously does more, but to find out what you have to read through a lot of material on the site, where a simple side-by-side comparison of features between versions, like this page showing the differences between Art Rage Lite and Art Rage 5 easily demonstrates what is or isn’t included.

I wonder if Dragon can tell the difference between pastor and pasta?

Bad design: Staples iOS app

The revised version of Staples’ iOS app lists products but no longer lets you know if a product is available both in-store and online or online-only. This matters when you use the app, find a product, then go to a store and are told, “lol, naw, we only have that online!”

To insure they are not wasting a trip to the store, a customer is forced to call ahead to check for stock, an inconvenience the app should eliminate, not create. This is bad design.

Best Buy’s app, on the other hand, not only tells you if a product is available in-store, it will provide a handy list of storers near you that have it in-stock. That’s good design.

Staples bad, Best Buy good.